2000 toddlers' emerging symbolic play: the influence of

University of WollongongResearch Online

University of Wollongong Thesis Collection University of Wollongong Thesis Collections

2000

Toddlers' emerging symbolic play: the influence ofpeers in the day-care contextHelen Sheila KowalskiUniversity of Wollongong

Research Online is the open access institutional repository for theUniversity of Wollongong. For further information contact ManagerRepository Services: [email protected].

Recommended CitationKowalski, Helen Sheila, Toddlers' emerging symbolic play: the influence of peers in the day-care context, Doctor of Philosophy thesis,Faculty of Education, University of Wollongong, 2000. http://ro.uow.edu.au/theses/1808

http://ro.uow.edu.au/


http://ro.uow.edu.au

http://ro.uow.edu.au/theses

http://ro.uow.edu.au/thesesuow



TODDLERS' EMERGING SYMBOLIC PLAY: THE INFLUENCE OF PEERS IN THE DAY-CARE CONTEXT

A thesis submitted in fulfilment of the requirements for the award of the degree

Doctor of Philosophy (PhD)

from

UNIVERSITY OF WOLLONGONG

by

HELEN SHEILA KOWALSKI, B.ED.ST., M.ED.

FACULTY OF EDUCATION

2000

THESIS CERTIFICATION

I, Helen Sheila Kowalski, declare that this thesis, submitted in fulfilment of the requirements for the award of Doctor of Philosophy, In the Department of Education, University of Wollongong, is wholly m y own work unless otherwise referenced or acknowledged. The document has not been submitted for qualifications at any other academic institution.

Signed:

Date: ?

Contents

BRIEF CONTENTS

Chapter Content Page

1 INTRODUCTION: THE NATURE AND SCOPE OF THE

ENQUIRY j

2 A REVIEW OF THE LITERATURE PERTINENT TO THE

CURRENT STUDY 17

3 THEORETICAL BACKGROUND 88

4 THE CONCEPT OF PLAY 115

5 METHODOLOGY 127

6 RESULTS 166

7 DISCUSSION OF RESULTS 214

8 CONCLUSION 261

BIBLIOGRAPHY 265

APPENDICES 279

l

Contents

TABLE OF CONTENTS


1 INTRODUCTION - THE NATURE AND SCOPE OF THIS ENQUIRY 1

1.1 THE PURPOSE OF THE ENQUIRY 1

1.2 THE CONTEXT IN WHICH THE PRESENT STUDY WAS SET 2

1.3 THE FOCUS OF THE STUDY 4

1.4 THE MAJOR HYPOTHESIS 5

1.4.1 The Dimensions of Play According to Westby (1991) 5

1.5 THEORETICAL SUPPORT FOR THE STUDY 6

1.6 RESEARCH INFORMING THIS ENQUIRY 7

1.6.1 Play as an Influence on Development 7

1.6.2 The Importance of Symbolic Play 8

1.6.3 Adult (Parental, Caregiver or Teacher) Influence on the

Development of Play 9

1.6.4 Peer Influence on the Development of Play 10

1.6.5 The Issue of Quality in Long-Day Child-Care 12

1.7 VARIABLES BEYOND THE CONTROL OF THE RESEARCHER 13

1.8 HYPOTHESIS AND ASSOCIATED QUESTIONS 14

1.9 THE NATURE OF THEMATIC PLAY 14

1.10 CONCLUSION 16

2 A REVIEW OF THE LITERATURE PERTINENT TO THE

CURRENT STUDY 17

2.1 INTRODUCTION 17

2.1.2 The Grouping of Children in Early-Childhood Settings 18

2.1.3 Current Practice in Australia 18

2.1.4 The Focus of the Current Investigation 19

2.2 LITERATURE EXAMINING OLDER PEER INFLUENCE O N PLAY 21

2.3 LITERATURE EXAMINING OLDER-PEER INFLUENCE O N PEER

INTERACTION 29

2.4 LITERATURE EXAMINING Y O U N G CHILDREN'S IMITATIVE

BEHAVIOUR 34

2.5 PROGRAMMING CONSIDERATIONS 43

2.6 LITERATURE EXAMINING S O M E ADULT INFLUENCES O N

TODDLERS'SYMBOLIC PLAY 45

li

Contents


2 (continued)

2.7 LITERATURE EXAMINING THE INFLUENCE OF OLDER

SIBLINGS ON THEIR YOUNGER SIBLINGS' PLAY 53

2.8 LITERATURE EXAMINING SOCIAL INTERACTIONS BETWEEN

SIBLINGS 62

2.9 LITERATURE EXAMINING SOME INFLUENCES OF ADULT

INTERACTION AND TEACHING STYLE 65

2.10 LITERATURE EXAMIMNG SOME INFLUENCES OF THE

DAY-CARE ENVIRONMENT ON TODDLERS' PLAY 72

2.11 LITERATURE EXAMINING SOME OF THE INFLUENCES OF THE

DAY-CARE ENVIRONMENT ON PEER INTERACTION 80

2.12 CONCLUSIONS 83

2.12.1 Choice of Early-Childhood Setting for Research Purposes 83

2.12.2 Observed Conditions 84

2.12.3 Length and Frequency of Observations 84

2.12.4 Programming for Mixed-Age Groups in Early-Childhood

Settings 85

2.12.5 Age Range of Children in Mixed-Age Groups in Early-

Childhood Settings 85

2.12.6 Selection of Older Peer Partner 85

2.12.7 Caregiver or Teacher Influence 86

2.12.8 Older-Peer Communicative Behaviour 86

2.12.9 Toddler-Participant Imitative Behaviour 87

3 THEORETICAL BACKGROUND 88

3.1 INTRODUCTION 88

3.2 THE AIM OF THE PRESENT STUDY 88

3.3 THEORISTS UPON WHOSE WORK THE STUDY IS BASED 89

3.4 THE COGNITIVE FUNCTION OF SYMBOLIC PLAY 92

3.4.1 A n Analysis of the Piagetian View of Play 95

3.4.1.1 Parallels Between Piaget's and Westby's View of

Symbolic Play 99

3.4.2 A n Analysis of the neo-Piagetian Perspective 101

3.4.3 A n Analysis of the Vygotskian View of Play 103

in


3 (continued)

3.4.3.1 Parallels Between Vygotsky's and Westby's View of

Symbolic Play 105

3.5 SYMBOLIC PLAY AS A FUNCTION OF COGNITIVE SOCIAL-

LEARNING THEORY 106

3.6 SYMBOLIC PLAY AS A FUNCTION OF PSYCHOANALYTIC AND

PSYCHO-SOCIAL THEORY 108

3.7 SYMBOLIC PLAY AS A FUNCTION OF ETHOLOGICAL THEORY 110

3.8 SYMBOLIC PLAY AS A FUNCTION OF ECOLOGICAL THEORY ... 112

3.9 CONCLUSION 113

4 THE CONCEPT OF PLAY 115

4.1 INTRODUCTION 115

4.2 THE HISTORICAL CONTEXT OF PLAY 115

4.3 DEFINITIONS OF PLAY 116

4.4 THE INFLUENCE OF PLAY ON DEVELOPMENT 117

4.5 SYMBOLIC (OR PRETEND) PLAY 117

4.5.1 Characteristics of Symbolic-Play Development 118

4.5.1.1 Dimension One: 'Decontextualisation' 118

4.5.1.2 Dimension Two: 'Thematic Content' 119

4.5.1.3 Dimension Three: 'Organisation of Themes' 119

4.5.1.4 Dimension Four: 'Self-Other Relationships' 120

4.6 SOCIAL P L A Y 120

4.6.1 Categories of Social Play 121

4.6.1.1 Unoccupied Play 121

4.6.1.2 Onlooker Play 121

4.6.1.3 Solitary Independent Play 121

4.6.1.4 Parallel Play 121

4.6.1.5 Associative Play 122

4.6.1.6 Cooperative Play 122

4.7 COGNITIVE P L A Y 122

4.7.1 Categories of Cognitive Play 123

4.7.1.1 Non-Play 123

4.7.1.2 Exploratory Play 123

IV

Contents


4 (continued)

4.7.1.3 Physical Play 123

4.7.1.4 Practice or Functional Play 124

4.7.1.5 Constructive Play 124

4.8 COGNITIVE EFFORT 125

4.9 THEMATIC UNITS OF PLAY 125

4.10 CONCLUSION 125

5 METHODOLOGY 127

5.1 INTRODUCTION: THE PURPOSE OF THE PRESENT RESEARCH ... 127


5.3 RESEARCH METHOD 128

5.3.1 Influences on the Research Design 128

5.3.2 The Research Method for the Present Project 129

5 3.3 Ethics Approval 131

5.3.4 Verification of Methodology and Instruments 131

5.4 SELECTION OF LDCCCs 132

5.4.1 Criteria for LDCCC Selection 132

5.4.2 The Quality Improvement and Accreditation System 133

5.4.3 The Effect of the Removal of Child-Care Subsidies 134

5.4.4 Features of the LDCCC Play Environment 135

5.5 SAMPLE 136

5.5 1 Recruitment of Toddler Participants 136

5.5.2 Characteristics of Toddler-Participant Groups 138

5.5.3 Recruitment of Older Preschool-Aged Peers 139

5.5.3.1 The Peabody Picture Vocabulary Test - Revised

(FormM) 140

5.5.3.2 Older-Peer Attendance at LDCCCS 141

5.6 PROCEDURES FOR DATA COLLECTION 143

5.6.1 Implications Arising from Changes in Commonwealth

Government Policy 143

5.6.2 Data Collection 144

5.6.2.1 Equipment 144

5 6.2 2 Video-Taped Observations 144

5.6.2.3 Length of Video-taped Observations 144

v

Contents


5 (continued)

5.6.2.4 Order of Observations 145

5.6.2.5 Conditions under which Participants were Observed

and Video-Taped 145

5.6.2.6 Provisions for Play in Condition One (Same-Age

Play) and Condition Two (Mixed-Age Play) 148

5.6.2.7 Provisions for Play in Condition Three (Dyad Play).. 148

5.6.2.8 Summary of Data Collection Completed in each

LDCCC 149

5.7 INSTRUMENTS FOR ADDITIONAL DATA COLLECTION 152

5.7.1 Parent Questionnaire 152

5.7.2 The LDCCC Environment 153

5.7.2.1 The Environment Rating Scale 153

5.7.2.2 Toys, Materials and Equipment 155

5.7.2.3 Ratios of Staff to Children 156

5.7.2.4 Staff Questionnaire 157

5.8 CODING A N D SCORING PROCEDURE 158

5.8.1.Inter-Observer Agreement 159

5.9 CODING FOR PLAY 159

5.9.1 Symbolic Play 159

5.9.2 Coding for Level of Cognitive Effort 160

5.9.3 Coding for Length of Thematic Unit 161

5.10 ADDITIONAL INFORMATION 161

5.10.1 Social Play 161

5.10.2 Cognitive Play 161

5.10.3 Toddler-Participant Imitative Behaviour 161

5.10.4 Older-Peer Communicative Behaviour 161

5 11 ANALYSIS OF DATA 162

5.11.1 Choice of Statistical Tests 162

5.11.2 Computer Software Packages 162

5.11.3 The Friedman Two-Way Analysis of Variance by Ranks 162

5.11.4 Wilcoxon' s Matched-Pairs Signed-Ranks Test 163

5.11.5 The Kruskal-Wallis One-Way Analysis of Variance by Rank 164

5.11.6 Logistic Regression Analysis 164

5.12 ANTICIPATED RESULTS 165

vi


5 (continued)

5.13 CONCLUSION 165

6 RESULTS 166

6.1 SUMMARY OF DATA COLLECTION PROCEDURE 166

6.2 CODING 168

6.2.1 Scoring Symbolic Play 168

6.2.2 Inter-Observer Agreement 169

6.3 STATISTICAL ANALYSIS 171


6.5 A PRELIMINARY EXPLORATION OF THE DATA 172

6.5.1 Descriptive Statistics: Means, Standard Deviations and

Medians for the Frequency of Symbolic Play Events 172

6.5.2 Descriptive Statistics: Means, Standard Deviations and

Medians for the Highest Levels of Symbolic Play 173

6.5.3 Descriptive Statistics: Means and Standard Deviations for the

Most-Frequently Occurring Levels of Symbolic Play 174

6.6 THE FREQUENCY OF SYMBOLIC-PLAY EVENTS 175

6.6.1 The Relationship Between Birth Order and the Frequency of

Symbolic-Play Events 176

6.6.2 The Relationship Between Age Group and the Frequency of

Symbolic-Play Events 177

6.7 THE HIGHEST LEVELS OF SYMBOLIC PLAY 178

6.7.1 The Relationship Between Birth Order and the Highest Level

of Symbolic Play 180

6.7.2 The Relationship Between Age Group and the Highest Level of

Symbolic Play 182

6.8 THE MOST-FREQUENTLY OCCURRING LEVELS OF SYMBOLIC

PLAY 184

6.8.1 The Relationship Between Birth Order and the Most-

Frequently Occurring Levels of Symbolic Play 184

6.8.2 The Relationship Between Age Group and the Most-


6.9 THE COMPLEXITY OF SYMBOLIC PLAY 188

6.9.1 Complexity and Participants' Ages 188

vii

Contents


6 (continued)

6.9.2 Explanatory Variables for Complexity 190

6.10 T H E RELATIONSHIP B E T W E E N G E N D E R A N D S Y M B O L I C

P L A Y 193

6.11 S U M M A R Y O F STATISTICAL RESULTS 193

6.11.1 Results using Friedman's Rank Test 193

6.11.2 Results using the Kruskal-Wallis Test 194

6.11.3 Results using Logistic Regression Analysis 195

6.12 ANALYSIS O F DESCRIPTIVE D A T A 196

6.12.1 Results of the Environment Rating Scale 197

6.12.2 Additional Components of the Long-Day Child-Care

Environment 199

6.12.2.1 Toys, Materials and Equipment 199

6.12.2.2 Staff-Child Ratios in LDCCCs 201

6.12.3 An Analysis of Responses to Parent Questionnaire 202

6.12.3.1 Responses to Question 1 and 2 202

6.12.3.2 Responses to Questions 3 and 4 205

6.12.3.3 Responses to Questions 5, 6, 7 and 8 207

6.12.3.4 The Parents in the Current Study 208

6.12.4 An Analysis of Responses to Staff Questionnaire 209

6.12.4.1 Response Rate 210

6.12.4.2 Analysis and Discussion of Responses 210

6.12.4.3 Staff Training and Qualifications 211

6.13 C O N C L U S I O N 212

7 DISCUSSION OF RESULTS 214

7.1 INTRODUCTION 214

7.2 THE HYPOTHESIS AND ASSOCIATED RESEARCH QUESTIONS ... 214

7.3 THE THEORETICAL BASES FOR THIS RESEARCH 215

7ASTATISTICAL ANALYSIS OF DATA 215

7.5 RESULTS FOR THE MAJOR HYPOTHESIS AND THE FIRST

RESEARCH QUESTION 216

Vlll

Contents


7.5.1 Frequency of Symbolic-Play Events 216

7.5.2 The Highest Levels of Symbolic Play 216

7.5.3 The Complexity of Symbolic-Play Activity 217

7.6 DISCUSSION FOR THE MAJOR HYPOTHESIS A N D QUESTION

ONE 217

7.7 LITERATURE SUPPORTING THE CURRENT FINDINGS 219

7.7.1 Research Featuring the Dyadic Model 219

7.7.2 Research Featuring the Group Model 220

7.8 THE RESULTS FROM A VYGOTSKIAN THEORETICAL

PERSPECTIVE 221

7.9 RESULTS FOR RESEARCH QUESTION T W O 222

7.9.1 The Relationship Between Birth Order and the Frequency of

Symbolic Play Events 223

7.9.2 The Relationship Between Birth Order and the Level of

Symbolic Play 223

7.10 DISCUSSION: THE RELATIONSHIP BETWEEN BIRTH ORDER

A N D SYMBOLIC PLAY 223

7.10.1 First-Boms 223

7.10.1.1 Relevant Literature 224

7.10.2Later-Borns 225

7.10.2.1 Relevant Literature 226

7.11 RESULTS FOR RESEARCH QUESTION THREE 226

7.11.1 The Participants' Weekly Attendance at LDCC 227

7.11.2 The Unsatisfactory Provision of Play Materials 227

7.12 CONCLUSIONS D R A W N F R O M THE RESULTS DISCUSSED

THUS FAR 228

7.13 ADDITIONAL RESULTS ARISING FROM THE STATISTICAL

ANALYSES 229

7.13.1 The Most-Frequently Occurring Levels of Symbolic Play 229

7.13.2 The Relationship Between Birth Order and the Most-


7.13.3 The Relationship Between Age and Symbolic Play 229

7.14 A C O M M E N T O N THE RESULTS FOR PARTICIPANTS

AGED 20 TO 22 MONTHS 231

IX

Contents


7 (continued)

7.15 THE RESULTS FROM A PIAGETIAN PERSPECTIVE 232

7.15.1 Symbolic Play 232

7.15.1.1 'Decontextualisation' (Object Substitution):

Dimension One 233

7.15.1.2 'Thematic Content': Dimension Two 234

7.15.1.3 'Organisation of Themes': Dimension Three 235

7.15.1.4 'Self-Other Relationships' (Decentration):

Dimension Four 235

7.15.2 Neo-Piagetian Perspectives 236

7.15.2.1 'Organisation of themes': Dimension Three from

the Neo-Piagetian Perspectives 236

7.16 THE RESULTS FROM OTHER THEORETICAL PERSPECTIVES ... 237

7.17 VARIABLES LIKELY TO H A V E INFLUENCED THE RESULTS .... 239

7.17.1 The Provision of Play Materials 239

7.17.2 Features of One Participating LDCCC 241

7.18 RESULTS INCONSISTENT WITH THE PREDICTIONS 243

7.18.1 Unsupported Predictions 243

7.18.2 Results Not Consistent with Predictions 244

7.19 DISCUSSION: THE CONTEXT OF THE CURRENT ENQUIRY 245

7.19.1 The Play Environment 245

7.19.1.1 Play Materials for Conditions One and Two (Same-

Age Play and Mixed-Age Play) 245

7.19.1.2 Play Materials for Condition Three (Dyad Play) 246

7.19.1.3 The Context of Mixed-Age Play (Condition Two)... 247

7.19.1.4 Group Numbers and Staff-Child Ratios 247

7.19.1.5 Access to Materials 248

7.20 IMPLICATIONS FOR EARLY-CHILDHOOD PRACTICE ARISING

FROM THE CURRENT STUDY 249

7.20.1 Mixed-Age Free Play 249

7.20.2 Time for Play 250

7.20.3 Space for Play 250

7.20.4 Toys, Materials and Equipment for Play 250

7.20.5 Adult Involvement in Symbolic Play 251

x

Contents


7 (continued)

7.21 DIRECTIONS F O R F U T U R E R E S E A R C H 252

7.21.1 The Influence of Adult Interaction and Teaching Styles 252

7.21.2 The Influence of the L D C C C Environment on Play and Peer

Interactions 253

7.21.3 The Influence of Older-Peer Communicative Behaviour 254

7.21.4 The Toddler Participants' Imitative Behaviour 255

7.21.5 Deferred Imitation 256

7.21.6 Benefits for Older Peers 257

7.21.6.1 Benefits for the Older Peers' Socio-Emotional

Development 257

7.21.6.2 Benefits for the Older Peers' Language and

Communication Skill Development 258

7.22 C O N C L U S I O N S 259

8 CONCLUSION 261

8.1 OVERVIEW OF STUDY AND MAJOR FINDINGS 261

8.2 OTHER INFLUENCES 262

8.3 THEORETICAL CONCLUSIONS 262

8.4 FINAL CONCLUSIONS 263

APPENDICES

Ai Excerpt from Westby, C. E. (1991). A scale for assessing children's pretend

play. In C. E. Schaefer, K. Gitlin, & A. Sandgrund

(Eds.), Play diagnosis and assessment. (131-161).

NewYork: John Wiley & Sons 281

Aii 'Expected Levels of Dimensions of Symbolic Play for Participants

According to their Age Group and the Westby (1991) Scale for Assessing

Children's Play 282

Bi Initial letter to Director of L D C C C 283

Bii Information to L D C C C staff 284

Biii Staff member consent form 285

Biv Letter of thanks to staff 286

Bv Staff questionnaire 287

Bvi Categories for specialised qualifications in Early-Childhood Education 288

xi

Contents

Content Page

Ci Parent information letter 289

Cii Parent consent form for child to participate in study 290

Ciii Letter of thanks to parents of participants 291

Civ Parent questionnaire 292

D Summary of PPVT-R (Form M ) scores for preschool-aged participants 294

E Environmental Rating Scale 295

Fi Play materials generally available in the outdoor play environment in

L D C C C s in Sydney, Australia 301

Fii Props supplied by researcher for dyad-play condition 302

Gi Play coding sheet 1 303

Gii Coding sheet for symbolic play 304

Giii Individual summary sheet for levels and dimensions of symbolic play in each

of the three conditions 305

Giv Older-peer communicative behaviour codes 307

G v Older-peer communicative behaviours by condition and type of

communicative behaviour 308

Gvi Frequency of each toddler participant's imitative behaviour in Condition

T w o and Condition Three 309

Hi Extracts from Minitab output for Friedman's Two-way Analysis of Variance

by Ranks 310

Hii Extracts from Minitab output for Kruskal-Wallis One-way Analysis of

Variance by Ranks 314

Hiii Explanation of Logistic Regression Analysis 321

Hiv Extract from S A S output for Logistic Regression Analysis 322

I Summary of statistical results 324

xii

Contents

LIST OF TABLES

Table Page

2.1. Literature Examining Older-Peer Influence on Play 22

2.2 Literature Examining Older-Peer Influence on Peer Interaction 30

2.3 Literature Examining Young Children's Imitative Behaviour 36

2.4 Literature Examining Some Influences of Parents on Toddlers'

Symbolic Play 47

2.5 Literature Examining the Influence of Older Siblings on their Younger

Siblings'Play 56

2.6 Literature Examining the Interactions Between Siblings 63

2.7 Literature Examining Some Influences of the Adult's Teaching and

Interaction Style 66

2.8 Literature Examining Some Influences of the Day-Care Environment

on Toddlers' Play 73

2.9 Literature Examining Some of the Influences of the Day-Care

Environment on Peer Interactions 81

3.1 Theoretical Approaches and their Relevance to the Current Study 90

5.1 Characteristics of Participating LDCCCs 134

5.2 Characteristics of Toddler Participants 137

5.3 Summary of Toddler-Participant Information 138

5.4 Characteristics of Toddler-Participant Groups 139

5.5 Characteristics of Preschool-Aged Participants 142

5.6 Summary of Older-Peer Participant Information 143

5.7 Play Materials Supplied for Dyad Play (Condition Three) 149

5.8 Summary of Total Data Collection 150

5.9 Summary of Complete Data Collection 151

6.1 Summary of Project Components 167

6.2 Inter-Observer Agreement 171

6.3 Frequencies of Symbolic-Play Events in Three Conditions 173

6.4 The Highest Levels in Each Dimension of Symbolic Play in Each

Condition (N = 24) 174

6.5 The Most-Frequently Occurring Levels in Each Dimension of Symbolic

Play in Each Condition (N = 24) 175

xiii

Table Page

6.6 The Frequency of Symbolic-Play Events in Three Conditions 176

6.7 The Relationship Between Birth Order and the Frequency of Symbolic-

Play Events in Three Conditions 177

6.8 The Relationship Between Age Group and the Frequency of Symbolic-

Play Events in each Dimension in Three Conditions 178

6.9 The Highest Levels of all Dimensions of Symbolic Play in Three

Conditions 179

6.10 The Relationship Between Birth Order and the Highest Level of

Symbolic Play in each Dimension in Three Conditions 181

6.11 The Relationship Between Age Group and the Highest Level of

Symbolic Play in each Dimension in Three Conditions 183

6.12 The Relationship Between Birth Order and the Most-Frequently

Occurring Levels of Symbolic Play in each Dimension in Three

Conditions 185

6.13 The Relationship Between Age Group and the Most-Frequently

Occurring Levels of Symbolic Play in each Dimension in Three

Conditions 187

6.14 Significance of Age in Logistic Regression of Complexity on Age for

Participants Aged 17-19 Months Exceeding Expected Level of Each

Dimension of Symbolic Play (according to Westby, 1991), in Each

Condition 189

6.15 Explanatory Variables for the Likelihood (or Odds Ratio) for

Complexity to Occur in Three Dimensions of Symbolic Play ('thematic

content', 'organisation of themes' and 'self-other relationships') 192

6.16 Environment Rating Scale Scores and Accreditation 198

6.17 Percentage of Total Occasions of Provision of Toys, Materials and

Equipment by Centre 201

6.18 Frequencies and Percentages of Parents' Responses to Questions 1 and 2 203

6.19 Frequencies and Percentages of Responses to Questions 3 and 4 206

6.20 Frequencies and Percentages of Responses to Questions 5 and 6 207

6.21 Frequency of Staff Responses by Training, Qualifications and by

Centre 210

6.22 Frequencies and Percentages of Staff Responses by Category of

Training and Qualifications and Number of Respondents 211

xiv

Contents

Table Page

6.23 Frequencies and Percentages of Positive Responses by Category 3 Staff

by Centre 212

xv

Contents

GLOSSARY In the literature concerned with the current research topic there exists some

variation in the use of terms. In this document the terms 'symbolic play' and 'pretend

play' are used synonymously. Long-day child-care centre is abbreviated to LDCCC

and similarly, long-day child-care to LDCC. The term 'same-age play' refers to a

group of similarly-aged children in a free-play situation and 'mixed-age play' indicates

a group of children aged from approximately 17 months to 60 months or more engaged

in free play. T w o children only, one toddler and an older, preschool-aged peer feature

in dyad play.

xvi

Abstract

ABSTRACT This research was to understand some of the important factors influencing the development of symbolic play in young children aged from 18 to 30 months. Most researchers and practitioners recognise play as developmentally beneficial in early childhood. Fisher (1992) in a meta-analysis of over 40 studies of children and play identified 'child-orientated' play - which includes symbolic or pretend play - as influential in enhancing young children's cognitive and social development. The current project, undertaken in long-day child-care centres investigated the influences of older peers on the development of symbolic play in young children. It was hypothesised that toddlers w h o engaged in free play with older peers would exhibit symbolic-play activity more frequently and at enhanced levels than when engaged in free play with their same-age peers.

Data collection comprised observations of children at play in: same-age outdoor free play, mixed-age outdoor free play, and dyad free play. The older peers' ages ranged from three and a half to approximately five years of age. Three, ten-minute observations of the participant in each condition were video-taped. In the dyad condition, the participant was paired with a familiar and 'self-chosen' older peer who was requested to help the younger child play with three different sets of equipment, each of which was likely to elicit symbolic play. Video-taped segments were rated every 30 seconds for the frequency and level of symbolic play in each of the four dimensions ('decontextualisation', 'thematic content', 'organisation of themes', and self-other relationships') identified by Westby (1991). The data gathered assisted the assessment of the level of the toddlers' symbolic play.

The parents of the toddler participants completed questionnaires providing information regarding: children's activities and interests; parental expectations of their child's developmental outcomes, and the socio-economic status of the family. Questionnaires completed by L D C C C staff members indicated the value they placed upon play and other components of the program; and the level of training and qualifications achieved in Early-Childhood Education by the respondent.

Results showed that toddlers engaged in free play in the dyad condition (Condition Three) displayed symbolic play more frequently and at higher, more complex levels than when engaged in same-age play or mixed-age play. A sibling effect was also found: toddlers who were 'first-borns' engaged in symbolic play more frequently and at higher, more complex levels in mixed-age play (Condition T w o ) than in either of the other two conditions.

It is suggested that older peers when engaged with younger toddler-aged children assume the role of the 'more-skilled other' (Vygotsky, 1987) providing 'guided participation' (Rogoff, 1990) for the toddler operating in their zone of proximal development. Additionally, dyad play is likely to support the mechanisms of 'joint attention' (Tomasello, 1995) or 'joint involvement' (Schaffer, 1996) considered as instrumental in human social cognition.

xvn

Acknowledgements

ACKNOWLEDGEMENTS

There are, as with most major goals undertaken in life, many people I wish to

thank - those without whose involvement this particular goal might never have been

achieved.

First, I wish to express m y great appreciation to m y three supervisors -

Associate-Professor Philip de Lacey, Dr Grace Masselos and Dr Shirley Wyver. Their

sharing of knowledge, expertise and time, together with the support and encouragement

they continually offered m e throughout the construction of this document was

inexhaustible.

I wish to acknowledge the parents and children, staff and administrators in the

long-day child-care centres in Sydney, Australia in which this research study was

undertaken. Without them this thesis would never have come to fruition!

M y thanks are due also to Nan Carter who as both tutor and friend, advised and

guided m e through the 'statistical' complexities.

I would like to thank m y parents, Joyce and Ken Baker, and m y husband,

Krzysztof Kowalski, first for their patience and secondly, for their unshakeable faith in

m y ability to complete this thesis.

Finally m y very grateful thanks to other family members, m y colleagues and

friends all of w h o m encouraged m e to continue when difficulties arose.

x v m

Chapter 1: Introduction

CHAPTER 1: INTRODUCTION - THE NATURE AND SCOPE OF THIS ENQUIRY

1.1 THE PURPOSE OF THE ENQUIRY

The play of young children and how it develops has attracted much interest and

research especially since the early 1970s (Johnson, 1983; Rubin, Fein, & Vandenberg,

1983). In early childhood settings, it is recognised that a curriculum centred on play is

both appropriate and supportive for young children's learning and development in the

early-childhood years (Schiller & Veale, 1989; Van Hoorn, Nourot, Scales, & Alward,

1993; Schiller, 2000).

Over the last thirty years research has increased into the various types of play

young children engage in and associated contributions to children's general

development (Fisher, 1992; Johnson, 1990). Of these types of play, symbolic play

(pretend or imaginative play) has been recognised as particularly influential in

children's cognitive development in that a number of intellectual skills are embedded in

such play (Johnson, 1976; Rubin et al, 1983). Symbolic play begins to emerge when

children are aged approximately 12 months - early 'toddlerhood' - as their skills in

deferred imitation and mental representation develop. Successful engagement in

symbolic play places a number of intellectual demands upon the player for specific play

behaviours. These behaviours are described as 'dimensions' of symbolic play by

Westby (1991) and comprise the following: 'decontextualisation', 'thematic content',

'organisation of themes' and 'self-other relationships'. (See Appendix Ai.)

Yet another field of enquiry has been the identification of influences in

children's environments that can favourably affect the development of symbolic play.

Researchers have explored a number of influences including context or environment,

adults, siblings, play tutoring, play training and free play (Fisher, 1992; Johnson, 1978).

It is the purpose of this enquiry to investigate the influence of older peers upon

toddlers' symbolic-play activity. More specifically this study investigates whether

toddlers, attending long-day child-care centres (LDCCCs) w h o experience free play

with older peers in either a mixed-aged group or dyad play, have a greater frequency of

symbolic-play episodes and exhibit higher levels of symbolic-play skills - earlier than

they do in homogeneous (same-age) groups.

Symbolic play allows children to incorporate a range of developing

representational abilities in their play. Object and situation transformations form the

1


core of symbolic activity. Symbolic activity involves the capacity to substitute one

object for another enabling that object to symbolically represent the required prop and

be used without the limitations of reality (Westby, 1991). One useful view of symbolic

play is that of Piaget (1962). H e postulated that symbolic play requires the child to use

'schemas' - a cognitive structure or 'format' for organising and making sense of

information experienced. Children adjust and modify their schemas as maturation

combines with increased experience and knowledge.

In terms of symbolic development, schemas arise first from the toddlers' own

experience, and take the form of single actions that are initially focused on the players

themselves. The complex symbolic play of the older preschooler features the adoption

of others' roles and the execution of ordered sequential actions as sophisticated - and

often 'fantastic' and improbable - themes are developed. The development of symbolic

play is an unfolding process closely linked with the child's increasing cognitive skills

and acquisition of language. As mentioned earlier, four discrete symbolic behaviours

(termed 'dimensions' by Westby, 1991) comprise symbolic play: Dimension One -

'decontextualisation' (object substitution); Dimension T w o - 'thematic content' (themes

or scripts evident in children's play); Dimension Three - 'organisation of themes'(the

coherence of children's themes and scripts); and Dimension Four - 'self-other

relationships' (children's ability to take roles or give roles to others). Each of these is

discussed in detail in Chapter Three (3.4.1.1 and 3.4.3.1).

1.2 THE CONTEXT IN WHICH THE PRESENT STUDY WAS SET

T w o areas of consideration prompt the current research question which is

concerned with cognitive outcomes in terms of the frequency and level of toddlers'

symbolic play. The first is a large body of research suggesting that younger children's

cognitive, social and language development benefits from these children's interactions

with their older peers; and the second is the programming practice of mixed-age

grouping adopted by many L D C C C s . Until recently in Sydney (the city in Australia in

which the study was undertaken) this practice was often implemented, for at least part

of the day, as a cost-effective method of staff deployment, and appears to have become

more widespread as a result of child-care centres' restructuring following the

withdrawal of Federal government subsidies in July 1997. Conversely, in some

L D C C C s multi-age or mixed-age grouping is a purposeful component of each centre's

2


educational philosophy consistent with a substantive amount of research (Goldman,

1981; Howes & Farver, 1987; Mounts & Roopnarine, 1987; Brownwell, 1990)

suggesting there are positive outcomes for grouping children in this manner.

The provision of long-day child-care (LDCC) has existed in Australia since the

early 1900s and by the 1920s what Spearritt described as "the educationally based

kindergarten and the child minding day nursery" (1979, p. 10) existed. However, as in

other industrialised societies, with the advent of sociological changes, experience in

L D C C C s has become an integral part of the lives of many children.

Changes have also occurred in family structures often resulting in the likelihood

of one parent alone being responsible for the majority of child-care: the increase in

nuclear families (together with mobility); the demise of the extended family

(traditionally a source of child-care support for many); and the changes in the patterns

of employment requiring family living patterns to accommodate both parents' ever-

expanding participation in the labour force, have all contributed to the family's

increased use of formal child-care settings.

Research investigating the effects of L D C C on children has burgeoned - as have

L D C C C s . Findings vary from those purporting to support child-care as enhancing

children's lives (Andersson, 1989; Field, 1994; Phillips, McCartney, & Scarr, 1987) to

those postulating the opposite, that time spent in collective care with other children

adversely affects the child (Schwarz, Strickland, & Krolick, 1974; Belsky, 1978, 1988;

Belsky, Garduque, & Hrncir, 1984). Concerns included the impact of non-maternal care

on attachment (Belsky, 1978, 1988; Belsky et al., 1984; N I C H D , 1996) to behaviour

that was "more aggressive, motorically more active and less cooperative" (Schwarz et

al., 1974, p. 502) exhibited by children who attended L D C C compared to participants

who had not experienced L D C C .

In Australia today more than 527,769 children are attending L D C C C s according

to the Australian Bureau of Statistics (ABS, 1999). Nine percent of children under the

age of one year attend L D C C C s but by the age of 12 months this increases to 24 percent

and by the age of four, 73 percent of young children are in L D C C C s or preschools

(ABS, 1999). Therefore a high percentage of toddlers are attending L D C C and are in

small-group situations, as opposed to family units, during much of the period in which

symbolic play emerges and develops.

It is common practice in L D C C C s to include mixed-age groups in their

programs. Such grouping entails toddlers aged from about 18 months to 30 months

3


sharing play materials, equipment and the physical and social environment, either

indoors or outside, with older preschool-aged children aged up to five or more years.

Although mixed-age grouping may be introduced into child-care settings as a result of a

centre's educational philosophy, grouping of this kind is often employed as a method of

combating staff shortages - a product of the centre's human resource logistics. In this

latter case mixed-age grouping is likely to occur either in the morning at the start of the

day or late in the day.

Following the withdrawal of Federal Government subsidies to child-care centres

in July 1997, L D C C C s were forced, at short notice, to reduce staffing and often

restructure the grouping of children to counteract the effects of reduced funding.

Consequently in the course of the present project, an increase in the grouping of

younger peers with older peers together for at least part of the early-childhood day was

seen to occur. Questions concerning outcomes with regard the impact of mixed-age

grouping on both the younger and the older child also seemed to become commonplace

in the period following subsidy withdrawal.

Historically, mixed-age sessions as those outlined above were often justified in

terms of cost-effectiveness and were known to cause some parents and caregivers

concern about their developmental appropriateness for the children involved.

Nevertheless in the past the practice of combining groups of younger and older peers

together was usually implemented after lengthy discussions and an exploration of the

implications of such practice with parents. However, the immediacy of the impact of

subsidy withdrawals was such that modifications and changes in L D C C C structures

resulted in a rapid reorganisation of staffing and grouping of children. Consequently,

opportunities for either parent consultation or parent participation and collaboration in

the decision-making process were limited. Further reference to this period of rapid

change in L D C C and its impact upon components of this project is made in the

subsequent chapters.

1.3 THE FOCUS OF THE STUDY

It seems, then, imperative that the focus remains on recent research suggesting

that concerns about mixed-age grouping might be unfounded and that developmental

benefits occur (Hartup 1983; Dunn 8c Dale, 1984; French, Wass, Stright, & Baker,

1986; Bailey, Burchinal, & McWilliam, 1993).

4


This study will examine one area of young children's cognitive development -

namely that of symbolic (pretend) play. It will seek to identify whether toddlers'

symbolic-play skills in the four dimensions of 'decontextualisation', 'thematic content',

'organisation of themes' and 'self-other relationships' as defined by Westby (1991) are

enhanced through experiencing free play with older, preschool-aged peers. T w o mixed-

age conditions are surveyed for this project: first toddlers engaged in outdoor free play

in a mixed-age group setting where preschool-aged peers were present, and secondly in

a dyad free-play situation whereby the toddler participant is paired with a familiar older

peer.

1.4 THE MAJOR HYPOTHESIS

The major hypothesis is that, toddlers engaged in free play with older peers will

exhibit symbolic-play skills in the four dimensions of symbolic play identified by

Westby (1991) - 'decontextualisation', 'thematic content', 'organisation of themes' and

'self-other relationships' - more frequently and at a higher level than when they are

engaged in free play with their same-age peers. (The four dimensions of symbolic play

according to Westby [1991] are explained below in 1.4.1.) Should the results of the

study support this proposition, then 'more-informed' programming could be

implemented furthering the provision of good quality care-giving for young children in

early-childhood settings.

1.4.1 The Dimensions of Symbolic Play According to Westby (1991)

Westby (1991) in developing a Scale for Assessing Children's Pretend Play (the

scale to be used in this project) identified four dimensions (or behaviours) of symbolic

play - 'decontextualisation', 'thematic content', 'organisation of themes' and 'self-other

relationships'. She postulated that, as young children's symbolic play is likely to follow

an even pattern of development; the skills in each symbolic-play dimension are likely to

be at a similar level. However, conversely, Westby (1991) recognised that variation in

dimension levels occurs, as each is linked to and reflective of children's different

cognitive, social and language abilities. (Westby's [1991] Scale for Assessing

Children's Pretend Play is discussed in detail in Chapter 4, 4.5 and also presented in

Appendix Ai.)

5


Level One in the 'decontextualisation' dimension remains appropriate

throughout the total age range of toddler participants (approximately 17 to 31 months)

in that play exhibiting more sophisticated or intricate symbolic activity in this

dimension is not expected within these age ranges. Although changes in the dimension

of 'decontextualisation' (according to Westby, 1991), occur when children are

approximately 36 months old it was decided to retain the dimension of

'decontextualisation' allowing toddlers' potential for realising levels beyond that

expected to be identified. It is also useful to note that the likely levels in the dimensions

of 'thematic content', 'organisation of themes' and 'self-other relationships' are

unchanged for those children aged 23-26 months. (This information is displayed in

Appendix Aii.)

1.5 THEORETICAL SUPPORT FOR THE STUDY

Current child-development theory in Early-Childhood Education reflects

Vygotskian propositions recognising the importance of the social context for young

children's learning and development. There exists now a wider recognition that others,

both adults and peers, w h o operate in young children's environments, play a crucial role

in the younger participant's acquisition of knowledge and skills. Schaffer refers to these

social interactions as "joint involvement episodes" (1996, p. 252) and describes other

people as being "perceived and progressively assimilated by the infant" (1996, p. 252).

These statements suggest two discrete components to the interactional process, the first

whereby the more-skilled other becomes a partner in the process aiding children to

make sense of the world, and the second relating to the role of that partner in shaping

the course of the child's development (Schaffer, 1996.)

This theoretical perspective, together with others seen as central to the study,

will be explored in more depth in Chapter Three.

6


1.6 RESEARCH INFORMING THIS ENQUIRY

Research relevant to the following five areas informs this enquiry:

- play as an influence on development;

- the importance of symbolic (pretend or imaginative) play;

- adult (parental, caregiver or teacher) influence on the development of

play;

- peer and sibling influence on the development of play;

- the issue of the quality of care provided in L D C C C s .

1.6.1. Play as an Influence on Development

The proposition that play is beneficial for young children in all developmental

domains is well supported. Forty years of research support such a proposition and in

that period many forms of play have been studied together with appraisal of various

outcomes. In 1992, Fisher undertook a meta-analysis of 46 studies involving play and

identified the category of 'child-orientated' play as being influential in terms of

positively affecting children's development. This category of play encompasses

imaginative play that, in keeping with symbolic play, requires the child to engage in

object and situation transformations. Such behaviour is manifested when readily

available objects are substituted for others that are unavailable. For example a block

might be used as a telephone receiver or a car or truck. Situational transformation

requires the suspension of reality in terms of the environment. A c o m m o n example is

the ready adoption of 'home corner' being used accordingly by the players.

Both the theorists Piaget and Vygotsky have contributed to the current

knowledge and understanding of symbolic or pretend play. Piaget believed that the

function of symbolic play "... is to assimilate reality to the ego while freeing the ego

from the demands of accommodation" (Piaget, 1962, p. 134). Vygotsky proposed that

symbolic play releases children from the constraints of reality. H e contended therefore

that symbolic play was an activity in which very young children under three years of

age are unlikely to be engaged due to situational constraints.

It is possible that young children attending L D C C are released from such

situational constraints and in fact experience an environment both rich in materials to

elicit imaginative responses and activity together with personnel w h o are actively

engaged in symbolic play. These factors m a y well support the emergence of toddlers'

symbolic play earlier than would otherwise be expected.

7


1.6.2. The Importance of Symbolic Play

Currently of particular interest to play researchers is the development and

subsequent benefits of thematic-pretend play - a complex form of symbolic play,

developed in the later preschool years, whereby children become involved in play

depicting activities and knowledge not related to their lives. Pretend play is defined by

Garvey (1990) as a subcategory of play requiring the child to represent actions, objects,

persons or places symbolically. Competency in symbolic play, involving the skills of

representation - initially using concrete props - is likely to be supportive in the

development of thematic pretend play in which the skills of 'concrete' representation

are replaced with those that are 'fantastic' and 'abstract'. (Thematic pretend play is

discussed further later in this chapter in 1.9.)

A second area of enquiry relevant to play is that of theory of mind - the

understanding of "what others feel, want and know" (Lillard & Curenton, 1999).

Dockett (1998) suggests that from both a Piagetian and Vygotskian theoretical

perspective shared pretend play is consonant with the development of theory of mind.

Cognitive dissonance (a Piagetian construct) and the joint construction of knowledge (a

Vygotskian construct) are each components of pretend play. Additionally, the influence

older siblings may have on younger siblings' development of theory of mind has been

the focus of research by Lewis, Freeman, Kyriakidou, Maridaki-Kassotaki, and Berridge

(1996) and Ruffman, Perner, Naito, Parkin, and Clements (1998).

Other writers in the field (Rubin, 1980; Vandenberg, 1981; Westby, 1991) view

the development of early symbolic play as the basis or springboard for the development

and promotion of several cognitive skills. Rubin (1980) postulates that cognitive skills

including one-to-one correspondence, conservation, decentration, seriation,

classification, reversibility, comprehension of transformations and identity together with

perspective taking, are among the abilities whose development is assisted by children's

engaging in pretend play. These skills or abilities are crucial for supporting children's

competencies required for not only understanding the physical world but also

understanding how and why the human inhabitants of that environment operate and

function as they do (Westby, 1991).

Symbolic play has been noted for its role in aiding the child to gather knowledge

about others (Smilansky, 1968). Mead (1932) recognised that symbolic play supported

the ability of the player to take on another's perspective, and was therefore instrumental

in the child's acquisition of self-concept and self-identity.

8


1.6.3 Adult (Parental, Caregiver or Teacher) Influence on the Development of Play

Winnicott - from a psycho-analytic perspective - suggests that early symbolic or

pretend play emerges as the child's relationship with the mother changes in accordance

with the infant's perception of himself or herself as a separate being as he or she attains

"unit status" (1965, p. 44). It would seem then of importance that children be supported

in the early development of these play skills. These skills have as their basis symbolic

play, which infants initially develop as a result of mother-child or caregiver-child

interaction early in life. Interestingly, O'Connell and Bretherton (1984) propose that the

child actively determines the effectiveness of the mother's or the caregiver's interactions

by heeding some of them and ignoring others. This suggests that the quality and

content of the subsequent symbolic play depends in part upon the sensitivity, and what

O'Connell and Bretherton describe as the ability of the adult play-partner to adopt the

strategy of "fine-tuning" to the play situation (1984, p. 363).

In a review of research, Fein and Fryer (1995) indicate that maternal

contributions to children's play, while perhaps positively affecting the amount of

symbolic play, have no long-term bearing on the quality of symbolic play. It is possible

to argue that the process of maternal play (it is likely that the mother will be the first co-

player in a child's life) enables the child, as Winnicott (1971) suggests, to recognise his

or her 'separateness'. Perhaps Bruner's notion of "tutor proneness" (1972, p. 697) is then

triggered allowing the child to be receptive to substantive input from interactions with

siblings, peers, family members, teachers and others in their social environment.

A substantial number of researchers have observed and commented upon the

caregiver's role in influencing the emergence of symbolic play in young children under

three years of age. Such commentaries range from a study, examining the impact of the

familiar context of mother-child interaction, undertaken twenty years ago (Johnson,

1978; Werner & Kaplan, 1963) to a more recent project, involving mothers'

participation as play partners in symbolic play, recorded by Haight and Miller (1992).

Mothers are commonly the caregivers in such studies, but more recently Farver and

Wimbarti (1995) and Haight, Parke, Ross, and Black (1997) have observed that fathers,

too, support the symbolic play of young children.

The method by which the adult play-partner supports the development of

symbolic play has been found by researchers to vary. For example, O'Connell and

Bretherton (1984) suggested that the mother's influence was indirect, that her presence

alone and/or the ability to respond, even if the response was not entirely appropriate,

9


might enhance children's symbolic play. Conversely, in their study of fathers as play

partners, Farver and Wimbarti (1995) found that fathers were keen to give explicit

guidance in supporting their children's play. A degree of consensus exists in emphasing

the importance of the caregiver-child interaction for symbolic play to successfully

occur.

1.6.4 Peer Influence on the Development of Play

Research into the effects of mixed-age grouping and any subsequent gains in

domain development has gained momentum over recent years. The notion that

interaction with older peers is important in terms of social and personality development,

has been supported by several researchers (Hartup, 1983; French et al., 1986). The

younger child, by observing and imitating the older peer, is able to acquire adaptive

social and cognitive patterns of behaviour such as role-taking skills which are a

component both of symbolic play and later thematic play. Possibly the most notable

research in this area was undertaken by Whiting and Whiting (1975) w h o noted that

older-aged peers were most instrumental in the nurturing and promotion of social

behaviours in their younger playmates, and employed the techniques of modelling and

reinforcement to encourage these desired behaviours. A study by Ellis, Rogoff, and

Cromer (1981) revealed that, given choice, as is available in mixed-age groups in

L D C C , children as young as two years old spend time with peers w h o are older or

younger by at least one year. Howes, Unger, and Seidner (1989) in their study of social

pretend play in toddlers concluded that being familiar or well acquainted with same-age

peers was likely to be instrumental in social-skill development, and resulted in quite

ritualistic interactions.

A n effect of L D C C could be that toddler interactions with familiar older-aged

peers may be likened to sibling interactions. Weisner refers to the c o m m o n experiences

of siblings resulting in a "complementarity of roles and relationships" (1989, p. 13)

appearing in their interactions and suggests this leads to greater competence for the

younger child when engaging in symbolic play.

The question arises whether toddler interactions with familiar older-aged peers

in L D C C could be compared to sibling interactions. Dunn and Dale in their study of

two-year-olds' pretend play in collaboration with both a sibling and mother described

sibling pretend play as the "close meshing of each partner's verbal and non-verbal

actions" (1984, p. 139). The question may then be posed whether children in full-time

10


child-care experience a 'sibling-type' relationship in that many c o m m o n experiences are

shared. It is likely that there would exist substantial social and emotional involvement

between children who have been together in day care for 'some time'. They have

experienced the same setting, may have developed c o m m o n expectations of each other

and would share a c o m m o n 'vocabulary of memories and past experiences' - not unlike

siblings.

A question which will be addressed in this study is whether or not the features of

sibling pretend play, identified by Dunn and Dale (1984), exist for children in mixed-

age group settings in L D C C C s . It may be possible that the participation by toddlers in

mixed-aged free play allows for increased symbolic-play opportunities with older peers

and subsequently leads to increased competence in symbolic-play skills.

A premise of Vygotskian theory, a theory in which this study is partially

grounded, is to recognise the role of the 'more-skilled and knowledgeable other' in

supporting the younger child's acquisition of skills and learning (1978). Vygotsky

stresses both the necessity, and the value in terms of transmission of culture and

institutional knowledge, of the social context for children operating in what he termed,

their 'zone of proximal development'. In keeping with these Vygotskian notions, the

older peer is the 'more-skilled other'. Therefore it is hypothesised that interactions

resulting from familiarity between toddlers and older-aged peers may result in enhanced

social-skill development, which may in turn enhance the development of play.

Younger children engaged in mixed-age groups were found to score higher than

those children in same-age groups across all developmental domains in the study

conducted by Bailey, Burchinal, and M c William (1993). The older children in the

group were identified as being competent models whose presence in the mixed-age

group created a more stimulating and exciting environment. The researchers found the

older child engaged in instructing, tutoring and teaching behaviours with the younger

children. Such behaviours are supportive of Vygotsky's proposition that a more-able

child can enhance a less-able child's experience and skill acquisition through joint

participation. Teachers setting up the environment for mixed-aged groups were found

to provide a greater variety of play materials, equipment and activities (Bailey et al.,

1993).

11


1.6.5 The Issue of Quality in Long-Day Child-Care Centres

In this study and others cited, the term 'quality' is used to describe an

environment in which the provision of good care-giving occurs - 'quality care'. An

environment described as low-quality or unsatisfactory indicates that the provision of

care is below that recognised as satisfactory or adequate.

Interest in outcomes for children attending L D C C C s persists. Of recent

consideration for researchers such as Andersson (1989) is the notion that children

classified as 'early entrants' to L D C C rate more favourably when assessed in subsequent

years than those children whose entry into L D C C was later. Investigations such as this

suggest the early development of symbolic play may be a favourable outcome in terms

of children's future developmental gains. It is relevant to note that Andersson's study

was conducted in Sweden where high-quality child care has been an outcome of a

concerted effort by the Swedish government and associated institutions.

In keeping with this research and other recently published work (Howes &

Smith, 1995; Howes, Phillips, & Whitebook, 1992), the quality of the care-giving

environment in the current project is recognised as an independent variable. A n

Environment Rating Scale is employed to control this variable, but it is recognised that

many inter-related factors influence the L D C C environment during the course of any

one day, from day to day and over time. Such factors include staff rosters which

determine the care-giving environment, the overall centre program including such

elements as use of indoor and outdoor space and the types and variety of toys, materials

and equipment provided for children's play.

In low-quality child care, higher caregiver-to-child-ratios often exist than in

high-quality child-care settings - a feature thought to be less conducive to the provision

of what is considered to be developmentally appropriate care for very young children

(Scarr, Eisenberg, & Deater-Deckard, 1994). Program planning is likely to be less

effective in low-quality child care with the consequence that activities provided for the

children m a y not reflect their developmental needs and interests or be as stimulating as

experiences provided in a high-quality child-care setting. A low-quality environment

tends to result in children spending large portions of their time in 'wandering' and

'uninvolved' activity, which in itself has implications for social development and

interaction, and consequently the development of symbolic play (Vandell & Powers,

1983, p. 498).

12


High-quality care appears to be associated with smaller caregiver-to-child ratios

and more appropriate, responsive, caregiver-child interactions. It is likely in such

settings that activities and materials suited to the child's stage of development will be

provided. These factors, together with the carer's appropriate expectations, are in

themselves instrumental in promoting social interaction among children.

Developmentally appropriate experiences and materials and the type of teacher-

child interactions, in L D C C , are categorised as 'process-quality indicators'. The

research findings of Howes et al. (1992) and Howes and Hamilton (1992a) suggest that

process quality aspects of the setting have a stronger relationship with child outcomes

than 'structural aspects' such as group sizes and teacher characteristics.

1.7 VARIABLES BEYOND THE CONTROL OF THE RESEARCHER

The need to identify a range of independent variables in terms of the older

preschool-aged peer or sibling is recognised. In their interactions with younger peers

(toddlers), older peers will exhibit differences in the amount of empathy expressed;

temperaments and ability will also vary. The nature of 'out-of-care' experiences is likely

to be different and behaviours engaged in at L D C C m a y also to some degree be

moderated by the older peers' positions in their family structure. The length of time the

older peers have attended the L D C C C may also affect their behaviours. The writer

accepts that these differences will exist and are, in the main, unable to be controlled.

Importantly, the older peers' receptive language will be assessed through the use

of the Peabody Picture Vocabulary Test-Revised (PPVT; Dunn & Dunn, 1981). The

PPVT-R is recognised as an appropriate instrument to use with preschoolers and is

"well-correlated with other measures of vocabulary" (Howes, 1997, p. 411).

Assessment of this kind will ensure that the older peers are within their expected range

of development at least in this regard.

13


1.8 HYPOTHESES A N D ASSOCIATED QUESTIONS

As stated on page 5, 1.4, the major hypothesis is then that toddlers engaged in

free play with older peers (in either mixed-age play or dyad play) will exhibit symbolic-

play skills in each of four dimensions of symbolic play identified by Westby (1991) -

'decontextualisation', 'thematic content', 'organisation of themes' and 'self-other

relationships' - more frequently and at a higher level, than when they are engaged in

free play with their same-age peers.

In conjunction with the major hypothesis, the following questions will also be

considered:

a) If the major hypothesis is supported, which setting - mixed-age play

or dyad play - is more influential on the frequency and level of toddlers'

symbolic play?

b) D o first-born toddler participants exhibit symbolic-play skills in each

of four dimensions of symbolic play identified by Westby (1991) -

'decontextualisation', 'thematic content', 'organisation of themes' and

'self-other relationships'- more frequently and at a higher level than

'later-born' participants?

c) Does the L D C C environment - which includes the quality of care

provided, the program of activities implemented, the materials and

equipment provided, staff training and qualifications, and other available

resources - support children's engagement in symbolic play?

1.9 THE NATURE OF THEMATIC PLAY

Symbolic play is a forerunner of thematic social play. Imagined events and

fictional narratives feature in thematic play. Wyver and Spence describe thematic

pretend play as "pretend play which deals with themes and events extremely remote

from the child's everyday life" (1995, p. 44). Their research findings suggest that as few

as 20 per cent of Sydney preschool children engage in thematic symbolic play. Such a

finding is of concern as children who are skilled thematic pretend players have been

found to be proficient in various areas of cognition (Wyver & Spence, 1999). It would

seem appropriate to alert early-childhood educators to the need for both close

monitoring of the development of symbolic-play skills and the promotion of such skill

development. One such avenue for the promotion of the acquisition of symbolic-play

14


skills, including thematic pretend play skills, may be through regular interaction with a

more-skilled or proficient other - for example an older peer who is both more adept at

symbolic play and incorporates more sophisticated themes.

Dunn and Dale (1984) have found that early thematic fantasy play is seen to

emerge in toddlers' symbolic play with siblings - as opposed to the domestic play,

which most often results from adult caregiver-toddler interaction. Farver and Wimbarti

(1995), in their study of fathers and children's symbolic play, observed that fathers

would suggest fantasy-play themes to their children. It would seem, then, of utmost

interest and importance to investigate the extent to which young children pretend with

various partners over the toddler (18-month to 30-month) period. Farver and Howes

(1993) concluded that young children engaged in more advanced forms of symbolic

play when they were pretending with a more-skilled partner than when they were alone.

This is in keeping with the Vygotskian (1962) notion of the child constructing

knowledge and developing new cognitive processes as a result of collaboration with

more competent others. Farver and Howes (1993) postulate that the scaffolding or

supporting of play by a more-skilled partner may be essential, and suggest that if such a

process is culture-specific then it is of paramount importance that the 'more-skilled

other' is proficient in the shared culture.)

That children voluntarily sustain involvement in symbolic play for longer

periods than other episodes featuring different types of play was noted by the Oxford

Preschool Research Group, headed by Jerome Bruner (Sylva, Roy, & Painter, 1980).

Their findings are pertinent to the present study. The Oxford Preschool Research Group

project (one of the earliest of its kind) was established in response to a huge expansion

of preschool education in Britain in the early 1970s. Its purpose was to evaluate the

benefits of preschool education in terms of children's cognitive, language and social

development. Symbolic play was recognised by the investigators as a type of play of

"only moderate complexity" (Sylva et al., 1980, p. 65) but requiring the child's

commitment and concentration. Bruner and his colleagues found that children engaged

in symbolic-play episodes for longer periods than those featuring other types of play.

Symbolic play often features a social element; however, currently, as previously

for the Oxford group, it remains almost impossible to identify whether it is the social

element or the dramatic content which attracts children and holds their attention.

15


1.10 CONCLUSION

The above discussion indicates that further investigation is needed in terms of

the effect of a L D C C setting on the development of toddlers' early symbolic play. The

present enquiry will examine how interaction with older preschool-aged peers in a

mixed-aged play group, and a dyad-play session, influences the evolution of such play.

The impact of the L D C C setting which encompasses the quality of care provided, the

program of activities implemented, the materials and equipment provided by staff and

other available resources, will also be examined. The question too of whether L D C C C

staff, trained in Early-Childhood Education, are more actively supportive of toddlers'

emerging symbolic-play skills than those who are not trained in Early-Childhood

Education will also be considered.

The current research, then, will seek to determine the answers to these two

questions: do toddlers w h o experience free play with older peers - either in groups or

dyads - exhibit symbolic-play skills more frequently, and do they do so at a higher level

than toddlers w h o are in same-age groups for free play?

To provide a basis for this enquiry a review of the relevant literature is presented

in the next chapter - Chapter Two. It is then necessary to consider the theoretical bases

arising from both the literature reviewed and the present enquiry. Several related

theoretical positions, particularly those describing cognitive, psychoanalytical, cognitive

social-learning, ethological and ecological approaches, are particularly relevant to the

proposed study. These will be described in Chapter Three.

The concept of play will be explained in Chapter Four, which also includes

detailed descriptions of the types or categories of play that are the focus of the current

project.

16

Chapter 2: Literature Review

i-

CHAPTER 2: A REVIEW OF THE LITERATURE PERTINENT TO

THE CURRENT STUDY

2.1 INTRODUCTION

Four main areas of literature inform the current research project on the influence

of older peers on the development of toddlers' symbolic play.

These are:

. older peer influences on both younger children's social interaction and

play;

. parental influences;

. birth order and sibling influences;

. the influence of the L D C C C care-giving environment including that of

the care-giving adults within that environment.

Literature arising from studies investigating each of these areas is examined in

this chapter.

Other factors considered as likely to be influential on young children's symbolic

play - the focus of this investigation - are contained within those four areas. For

example, the older peers' facilitative behaviour toward the younger peers and the

younger peers' imitative behaviour frequently feature in the research concerned with

mixed-age grouping. Facilitation and imitation are also components of sibling and

caregiver interactions. Components of the L D C C C environment, including the quality

of provision of care and staff training and qualifications, are considered in the literature

presented in the final part of this chapter (2.9, 2.10 & 2.11).

Mechanisms accounting for the differences predicted between participants'

levels of symbolic activity are likely to include both the older peers' communicative

behaviours and the younger peers' imitative behaviours. Therefore specific references

are made to these behavioural components as they arise in the literature reviewed. In

addition, due to the centrality of imitation in the emergence of early symbolic play

(Tomasello, Savage-Rumbaugh, & Kruger, 1993; Tomasello, 1995; D u n h a m & Moore,

1995), a number of studies featuring imitation as the dependent variable - and thought

to be useful to the current investigation - are presented in part 2.4 of this chapter.

17


2.1.2 The Grouping of Children in Early-Childhood Settings

Three significant and distinct approaches to grouping children in practice (and

subsequently research) in early-childhood settings emerged in the 1970s. Until that

time, children in L D C C C s , preschools and primary-school classes were most often

grouped with their same-age peers. Conversely though, mixed-age groups were the

norm in Family Day Care - where (in N e w South Wales, Australia) a maximum of five

children under school-age (including the carer's own children) were cared for at any

one time. Mixed-age groups were known to exist (though not commonly) in other

early-childhood settings. The third approach - grouping children for part of the day

only, in mixed-age cohorts - is now more widely practised in a range of early-childhood

settings.

The first group of studies examined are those featuring young children's

engagement in activities with older, more accomplished peers. The foci of these studies

cover a range of outcomes relating to aspects of play development (most pertinent to the

current study) and areas of cognitive development and social development. References

to older-peer behaviour and younger-peer imitation are often contained within such

studies.

Research conducted in settings other than L D C C C s has been referred to in this

section for two reasons. The first is to illustrate the breadth of interest existing in the

outcomes of mixed-age grouping across a range of age groups and early-childhood

settings, and the second to contextualise the current study

2.1.3 Current Practice in Australia

Mixed-age grouping occurs in Australia for pragmatic reasons however, children

attending early-childhood settings are, where numbers permit, generally grouped

according to age - in keeping with the once widely held beliefs that young children's

development and learning benefit from experiences with their same-age peers. This

practice, the grouping of children by age, was noted by Austin (1976) as widespread

throughout the world. H e commented that "in most countries, children are grouped by

single year of age" (1976, p. 48). Interestingly, Austin also reported on the then

somewhat experimental activity occurring in Sweden where the practice of 'sibling

groups' was gaining momentum. The age span of children in these groups was a

minimum of three years to as much as six years (Austin, 1976; Sundell, 1994) - hence

18


classes or groups of children represented a six- to seven-year age span in the same

caring and instructional environment.

Concurrently, in contexts where it was impractical to group children in same-age

groups due to either low numbers of children attending the institution or their ages

spanning a wide range, children were found to be successfully developing and learning

in an environment catering for a mixed-age range. The Family Day Care setting was

one such environment where small numbers of children attend in 'family-like' age

groups furnishing opportunities for siblings to operate in a shared early-childhood

setting. Such settings catered for the research interests of Rothstein-Fisch and Howes in

1988 and more recently Dunn, Kontos, and Potter in 1996.

Simultaneously research questions arose as to the benefits or otherwise of young

children sharing learning contexts with both older and younger peers or siblings in the

more formal environments of primary or elementary schools. Swedish researchers

supplied a large quantity of research findings in this category. In both Swedish

elementary schools and early-childhood settings catering for children of younger than

school age, children were placed in 'sibling groups' (Sundell, 1994).

Research into mixed-age grouping in either preschools or L D C C C s continues to

gather momentum. A reasonable body of research reporting on the perceived

advantages and disadvantages of mixed-age and same-age grouping currently exists.

Such research focuses on a number of domains of young children's development and

learning, including play behaviour (Lougee, Grueneich, & Hartup, 1977; Goldman,

1981; Howes & Farver, 1987; Blasco, Bailey, & Burchinal, 1993; Lyytinen, 1995).

2.1.4 The Focus of the Current Investigation

The influence of older peers on the development of toddlers' symbolic play

skills is the focus of the current study. It draws upon the findings of the research

reviewed in this chapter that has investigated a range of outcomes arising from the

practice of mixed-age grouping in the preschool years. The research reviewed includes

those investigations focusing on either social or cognitive aspects of young children's

development, as each is a component of symbolic play.

A range of aspects of play, including symbolic (or pretend) play, feature as the

dependent variable in the studies reviewed - each examining the benefits of mixed-age

grouping. Importantly, none include the fine-grained developmental analysis afforded

by a scale such as Westby's (1991) Scale for Assessing Children's Pretend Play.

19


The settings for the studies reviewed in this section include Family Day Care,

Long-Day Child-Care, preschool and early primary-school classes. To date, research

into mixed-age grouping has received little attention in Australia, so that the literature

reviewed for this study is mostly sourced from the United States. Studies from Sweden

and Italy are also cited.

It is important, when reviewing the research describing the influential features of

child-care on children's learning and development, to keep in mind that the

characteristics of L D C C settings, in Australia, vary a great deal. In fact there exists

great variability from one L D C C C to another. For example differences m a y be found

in: the kind of play materials, equipment and toys provided; the programming and

organisation of space; the allocation of time for play; the opportunities for free play and

structured play; and the training levels of staff. (The last difference may range from no

training at all to the possession of tertiary degrees in Early-Childhood Education.) The

length of the child-care day makes it difficult to control for these differences when

undertaking research in children's day-care social and physical environments.

To ensure ecological validity, this present project was conducted under semi-

naturalistic conditions in a range of L D C C C s selected as a representative cross-section

of those in Sydney implementing mixed-age grouping as a regular component of their

daily program.

Australian children may experience a number of different child-care contexts

during any week - the L D C C C may be only one of these experiences (Wangmann,

1995). In 1999, 60 percent of children (439,600) regularly used L D C C for 5-19 hours a

week while three percent attended for 45 hours or more per week (ABS, 1999). It is

possible that for many children part of their week was also spent in the care of their

parents, relatives or in some other form of informal care, for example, neighbours,

friends or babysitters.

In conclusion, this section of the literature review considers the research first

examining older peer influence on younger peers' development in a range of play

behaviours including: social-pretend (or symbolic) play, play mastery, symbolic play;

and secondly peer social interaction.

20


2.2 L I T E R A T U R E E X A M I N I N G O L D E R - P E E R I N F L U E N C E O N P L A Y

Studies relevant to this section - those examining peer influences on the

development of social-pretend (or symbolic) play, play mastery and symbolic play - are

listed in Table 2.1. As in all Tables in this chapter, details listed include: author(s), date

of publication of study, the early-childhood setting in which study was conducted, the

number and age (in months) of participants, the number and length of observations

made, the independent variable and finally, the measures used. A brief summary of the

research findings relevant to the current study is also incorporated in each Table.

21

CO

•50 3

1 E

CO

4)

3 CO

ea 4)

s 'S > 3 4) 3 4)

a o

a

CO

S5-8

4) i-p

00*3

Z < E 00 3

ts 4> 00

Author(s)

Date

Younger peer exhibited more cooperative

social pretend play when paired with a

five-year-old peer.

Social-pretend play

(Howes, 1980)

Social behaviour (adapted from Giffen,

1984)

Same-age and

mixed-age dyads

1 x 15 min.

(same-age)

1 x 15 min.

(mixed-age)

SO -tf

U U U

9

Howes & Farver

(1987)

(Study 1)

Younger peer exhibited more cooperative

social pretend play and imitation when

paired with a five-year-old peer

Social-pretend play

(Howes, 1980)

Social behaviour (adapted from Giffen,

1984)

>̂ T3 U 60

ca -a 4) X

3 x 15 min.

(2 x same-age,

1 x mixed-

age)

SO -tf

u u u

s

Howes & Farver

(1987)

Study 2

Younger children engaged

in more

mature forms

of play when in mixed-age

groups.

Social play (Parten, 1932)

Cognitive play (Smilansky, 1968)

3 O

•x '3 00 O

u o 00

3

>o X o y-t

18

33-44

a to 4)

Mounts &

Roopnarine

(1987)

Mature

levels of play observed

in both

settings,

little age segregation

in mixed-

age groups.

Nested

social-cognitive scheme

(Rubin

et al., 1976)

Social and cognitive

modes of play

94

33-36

o o 4=

o CO

s Qy

Roopnarine,

Ahmeduzzaman,

Donnely,

Gill,

Mennis, & Arky,

(1992)

CO

00 3

'•£ 3

E

CO

3 CO

ea 4)

4>

"S > 3 4) T3

§ 4>

-a 3

p — 1

CO

z-8

«•. CO 4) r-p

0 0 «

Z < E oo 3

00

Author(s)

Date

Developmental maturation is required for

social mastery.

Non-mastery, pre-mastery, mastery,

social mastery (adapted from Purdue

Free Play Mastery Motivation, Wachs,

1987b)

Social mastery

in

play

3

WN

X -tf

48

12-48

U V

u

a

Blasco, Bailey, &

Burchinal

(1993)

Higher scores in

all developmental

domains achieved by younger

children in

mixed-age groups.

Battelle Developmental Inventory

(Newborg et al., 1984)

•UP

3 4>

E CL,

13 > 4) T3

13 fe > O 3

vn

X -tf

59

21-67

U U

u Q yy

Bailey, Burchinal,

& McWilliam

(1993)

Increased play activity, increased

more complex symbolic play in the dyad.

rt 00 ON *—P C

o CO

3 4) y\

o • >

ea 4>

ca

Pretend play

p — i

X -tf

81

24-72

LDCCC

but aside

from

regular

program

Lyytinen

(1995)

Younger peers showed more complex

cognitive play behaviour with older

peers.

Peer

play (Howes, 1980)

PPVT-R (Dunn & Dunn, 1981)

Cognitive play (adapted from

Smilansky, 1968)

Play with objects (adapted from

Howes & Stewart, 1987)

Social, cognitive and

language

development

4 hours over 3

visits

57

30-48

U

e

Dunn, Kontos, &

Potter

(1996)


A chronological starting point for discussing research related to both mixed-age

groups and symbolic play is the writing of Peller (1954, 1978) who, from a

psychoanalytic approach, recognised the value and role of symbolic play as a cathartic

activity. A base line for research investigating the benefits of grouping differently aged

peers for play is also furnished by Peller (1978). More than forty years ago she

recognised the many benefits of mixed-age interaction - mainly extolling the benefits

accrued by the older peer (1954). Nevertheless, Peller (1978) alerted practitioners to the

advantages of such grouping.

The social-pretend play of two-year-olds paired with same-age peers and later

with five-year-olds was the focus of two related studies by Howes and Farver (1987).

The method for each study required observation of two dyads - one same-age and the

other a mixed-aged dyad. All dyads were same-sex. In the first study each dyad was

observed once only for 15 minutes. Howes and Farver (1987) recognised that the

younger age group (the two-year-olds) exhibited more cooperative social-pretend (or

symbolic) play when paired with five-year-old partners.

In the second study, data were collected from three observation sessions. The

first observation was of children in their same-age dyad. The second was the

observation of the mixed-age dyad in which the five-year-old partner was requested to

'teach1 the younger child to play with the materials provided. Same-age dyads were then

observed again to complete the third and final observation. Toddler participants in the

second condition (where the older peer was requested to 'teach' the younger child)

engaged in a significant amount of imitative behaviour and engaged in more simple and

cooperative pretend (or symbolic) play than toddlers in the control group (Howes &

Farver, 1987).

The second study undertaken by Howes and Farver (1987) revealed that toddler

participants did not transfer to same-age play the more complex play patterns they

engaged in with the older peers. It transpired, that toddlers were more instrumental in

constructing simple social-pretend (or symbolic) play (where each engaged in pretence

actions), with same-age peers than with older children. This finding suggests that the

toddler participants were more effective at cooperating with children of their own age.

Toddlers, when interacting with other children of similar age, supported their

constructive-pretend (or symbolic) play verbally through metacommunications, giving

directions to the play and directly instructing behaviours. These results would suggest

that one session of mixed-age play alone is not sufficient to change toddlers' play

24


patterns, but the results also raise the question whether regular extended opportunities

for toddlers to engage in mixed-age play may be effective in enhancing their social play

activity.

The social cognition of children aged from 42 to 54 months was the dependent

variable in a study reported on by Mounts and Roopnarine and published in 1987.

Again same-age and mixed-age groups featured as the independent variable. Children

were observed for 10 periods of five minutes in free-play periods, and the frequency of

a range of interactive behaviours was recorded. The researchers reported that the

younger child's interactive-play behaviour in the mixed-age group was more mature

than that exhibited by those participants in the same-age group. Three-year-olds

participating in the mixed-age group were more likely to engage in a range of

constructive play activities including social-constructive play.

Mounts and Roopnarine (1987) were also interested in the themes and content of

children's dramatic play - a form of play reflecting imaginative or symbolic content.

(Dramatic play is the term Smilansky [1968] used to describe symbolic play.) A broad

range of dramatic-play themes was observed in both settings. However, no significant

differences were found for either the themes or content of dramatic play engaged in by

the younger children in the mixed-age group than those displayed by participants in the

same-age group.

Observations made over a six-month period were central to a study by

Roopnarine et al. (1992) exploring social and cognitive aspects of play. Their three-

and four-year-old preschool participants were grouped in same-age or mixed-age

groups. Data were collected by observers using checklists of play activities and play

partners during a total of 12, five-minute observation periods. The 12 observations

were taken on two occasions, six months apart. While no formal preschool selection

process was implemented, care was taken to conduct the study in centres where program

philosophy, teacher training and beliefs, and the provision of toys, play materials and

equipment were similar. Findings in favour of mixed-age grouping were varied and

inconclusive. Using the play scale constructed by Rubin, Maioni and Hornung (1976)

in which Parten's (1932) social play categories and Smilansky's (1968) cognitive play

categories were combined, Roopnarine et al. found cooperative-dramatic play featured

more often in the three-year-old same-age groups while cooperative-constructive play

occurred more frequently in the mixed-age group. Three-year-olds grouped with age-

mates engaged in parallel-dramatic play, but when grouped with four-year-olds

25


displayed more solitary-manipulative play. Roopnarine et al. suggest that the mixed-

age group differences in children's language and cognitive ability may hamper shared

cooperative symbolic or dramatic play.

The findings of a pilot study, undertaken five years earlier by two members of

their research team, encouraged Blasco et al. (1993) to implement a study focusing on

the mastery behaviour of young children. (The earlier pilot study indicated that mixed-

age conditions promoted an increase in the cognitive content of the toddlers' play.)

Social mastery, defined as social and dramatic (or symbolic) play in free-play

situations, was subsequently investigated, by these three researchers, as one of a range

of mastery dimensions. The hypothesis proposed by Blasco et al. (1993) was that social

mastery, classified as a combination of social and dramatic (symbolic) play would vary

in the participants as a function of the age and stage of development of their peers.

Participants in the study were aged from 12 to 48 months and were randomly assigned

in either same-age or mixed-age groups. One component of interest in the work of

Blasco et al. is the notion of play mastery. This term is defined as the extension of skills

already acquired by children, a result of imitation and modelling - a construct in

keeping with both Burner's (1973) and Bandura's (1977) notion of children imitating

models for learning new skills and behaviours.

The results achieved by Blasco et al. (1993) are interesting in that significant

effects were registered for social mastery but these effects did not appear to be related to

the type of group, either mixed-age or same-age, in which children were placed. Blasco

et al. postulated that the lack of difference between children's achievement in play

mastery were more related to developmental maturation than the effect of mixed-age

grouping. (In the current study, the toddler participants interact with their older peers in

two conditions - one is a mixed-age free-play session and the other is dyad free-play

period. It is expected that the results of this present study will assist in clarifying the

condition in which older peers might influence younger peers play activity.)

Assessment across a number of developmental domains was the focus for a

project conducted by Bailey, Burchinal, and McWilliam and reported on in 1993.

Theirs was a longitudinal, four-year study of children aged initially between 21 and 67

months placed in either same-age or mixed-age groups. Again the independent variable

was same-age and mixed-age groups while the dependent variable was children's

developmental status, measured by The Battelle Developmental Inventory (BDI). This

process entailed the scoring of more than 300 items to assess the participants' status

26


across five domains of development: personal-social, adaptive, motor, communication

and cognitive. The younger children in the mixed-age groups were seen to achieve

higher scores, across all developmental domains, than those children in the same-age

group. However, for the four-to-five-year-old children in the study, those

developmental benefits were not maintained. A finding that Bailey et al. (1993) suggest

may indicate that older preschool-aged children may experience additional benefits

beyond those found in same-age groups.

It is possible, in research projects such as those of Bailey et al. (1993) and

Bailey, McWilliam, Ware and Burchinal (1993), that other variables might be

identified. For example, as postulated by those researchers, both the nature and the

variety of materials supplied by staff for mixed-age grouping m a y be greater than those

provided for same-age grouping activities. It is likely that materials supplied for a

mixed-age group would accommodate the expected higher skill level of the older more

mature player. In early-childhood settings it is sound practice to furnish the

environment with developmentally appropriate toys, equipment and materials. The

early-childhood practitioner, a professional trained in child development and

recognising the nature of children's learning, is likely to provide play materials first to

accommodate children's emerging skills and second to nurture and promote further skill

development. Kaplan and Conn reported that "great enrichment, organisation and

personalisation", (1984, p. 91) was evident in the L D C C C after staff attended a brief

training period.

As described earlier in this thesis, young children's symbolic play develops from

very simple representation using realistic props (sometimes referred to as 'high-structure

toys') to the complex representation exhibited by older children whose ability to engage

in object transformation is supported by the provision of 'unstructured' materials (Van

Hoorn, Noorot, Scales, and Alward, 1993, p. 174). If both sets of equipment are

available to support mixed-age grouping then younger children have access to more

sophisticated materials or toys earlier than m a y be likely in a setting catering for

children of a similar age.

In 1995, Lyytinen examined the symbolic play of children, aged between two

and six years, in two conditions. Lyytinen was interested in the effect or function of age

and play situation (solitary or dyadic) on three components of symbolic play:

decentration; decontextualisation; and integration. (These components are synonymous

with Westby's [1991] dimensions of symbolic play.) Participants were video-taped for

27


at least eight minutes while engaged in solitary play and dyadic play - the partner in the

latter condition being a familiar same-age, same-gender peer. Play materials provided

were such that should elicit a range of symbolic play. Lyytinen (1995) categorised play

as non-symbolic, transitional or decentred.

Dyadic play resulted in less non-symbolic play at all ages. Importantly, for the

current research, all age groups exhibited decentration - the ability of players to act as

others or involve others in their symbolic play (Westby, 1991) under both conditions.

In addition, children in all age groups exhibited more decentration when they were each

paired with another child.

Decontextualisation, the process by which pretend play becomes less dependent

on realistic props, also occurred more in the dyad condition. Lyytinen's (1995) findings

lend support to the current hypothesis that more sophisticated pretend play will occur

more frequently as a result of interactions with other peers, especially older peers.

Support for mixed-age grouping as influential in enhancing younger children's

development in a number of areas comes from a study conducted by Dunn, Kontos, and

Potter (1996) in Family Day Care. The average group size was seven with participants'

ages ranging from 30 to 48 months. In the analysis of 10-second time samples during

four, five-minute observations, younger children's cognitive play activity was found to

be more complex when they were engaged with older peers.

In summary, the literature reviewed supports the practice of mixed-age grouping

as a method of enhancing the development of younger children's play skills. Dunn et

al. (1996) in keeping with other researchers (Roopnarine et al., 1992; Sundell, 1994)

caution that while interactions with older peers are conducive to the younger

participants' development, less-competent behaviours, including play, m a y be exhibited

by the older peers. Peller (1978) however, identified a number of benefits for the older

peer arising from engagement with less-skilled peers.

In conclusion the findings from studies of mixed-age grouping are somewhat

equivocal and prompt the current study. The methodology in place for this investigation

presents the toddler participants with two opportunities - mixed-age play and dyad play

- to engage in free play with their older peers. It is expected that the results of this

present study will assist in clarifying the condition in which older peers might influence

younger peers' play activity.

Nevertheless, the studies reviewed are perhaps indicative of the importance of

providing opportunities for children to engage in a range of age-groups, each reflecting

28


developmentally appropriate materials and programming, throughout the early-

childhood day.

2.3 LITERATURE EXAMINING OLDER-PEER INFLUENCE ON PEER

INTERACTION

A commentary from Hartup (1976), taking an ethological theoretical stance

together with a cultural perspective, provides a useful starting point in the range of

literature to be reviewed here. Hartup's premise was that in evolutionary terms mixed-

age groups, for both humans and other primates, appeared to have existed longer than

same-age groups and that learning and development were likely to have occurred in

small social contexts. With the arrival of industrialisation in conjunction with massive

increases in the population, compulsory schooling children and latterly an extensive use

of L D C C C s , children were grouped, usually for ease of management, in same-age

groups. Therefore same-age groups have evolved quite recently.

This part of the chapter reviews the literature focusing on a range of studies

using the independent variables of mixed-age or dyadic situation to examine the

outcomes on young children's peer interactions. Table 2.2 contains the details of the

most informative of these studies - discussion follows the Table.

29

CO

00 3 3

E

& to

ea 4>

4)

'5 > 3 4> T3 3 4)

a. 4) T3 3 M CO -O

o fc

5o-S fc < E

00 3

•a CO

Author(s)

Date

Social and behavioural adjustments and

accommodation were made by each

partner

in the mixed-age group.

An increase in the younger

peer's attention to

the play

activity in the mixed-age dyad.

Positive social interactions

Negative

social interactions (Author's

scale)

Verbal

interaction system (Garvey &

Hogan, 1973)

Same-age and

mixed-age groups

3

•g O X CN

54

38-62

o o JS CJ CO

8 OH

Lougee,

Grueneich,

& Hartup

(1977)

Younger peers spent

less time

in parallel

play.


Behaviour observation guidelines

(Blurton Jones, 1972; McGrew, 1972)

Same-age and

mixed-age groups

10x3

min..

116

33-44

o o -3

o CO

4)

Goldman

(1981)

Parti.

Choice

of playmate appeared to be influenced

by sex.


Behaviour observation guidelines

(Blurton Jones, 1972; McGrew, 1972)

Same-age and

mixed-age groups

3

"S m X

o yy

116

33-44

o o JS

u CO 4) UP

AH

Goldman

(1981)

Part 2.

Two-year-olds were preferred peers.

Toddlers more active and aroused during

complex peer interaction.

Peer

play (Howes, 1980)

Social behaviour (Authors'

scale)

CO

§-O

feb 4) 00 ca i

73 4) X 3

X NO

30

16-23

Rothstein-

Fisch & Howes

(1988)

CO

00 3

• -i-t

•s yy

CO

4)

CO

ea 4)

s 4)

>

1 1 CL

I? u^

CO

Xi O

fc

oo'S

fc < E 00 3

'£t 4) CO

Author(s)

Date

Toddlers

adjusted social behaviour (both

content and complexity) to that of older

peer.

Peer-directed behaviour (adapted from

Eckerman et al., 1975; Lamb, 1975; and

Lewis et al., 1975)

Same-age and mixed-age

dyads

1x15

min.

62

18-25

U

u u Q yy

Brownwell

(1990)

Younger peers spent 67% of their time

playing with older peers

in mixed-age

groups.


BDI (Newborg et al., 1984)

Style of play (Authors' scale)

Same-age and mixed-age

groups

4x15

min.

32

18-60

u u u Q

Bailey, McWilliam,

Ware, & Burchinal

(1993)


Dyad play with a same-age peer and an older peer was the focus of Hartup's

research with colleagues Lougee and Grueneich (Lougee, Grueneich, & Hartup, 1977).

Participants were three years old and were observed twice, once in each condition,

engaged in free play.

Results were perhaps a little inconclusive. First, in the mixed-age dyad, younger

children attended more to the play activity than they did in the same-age dyad condition.

In addition, in the mixed-age condition each of the younger peers looked less frequently

at the older peer play partner than they did at their partners in the same-age play

situation. However, the surrounding environment captured less of the younger peer's

attention in the mixed-age dyad situation. Lougee et al. (1977) also noted that each

member of the mixed-age dyad was able to make both social and behavioural

adjustments in their interactions as they each accommodated the demands of the dyad

partner. The investigators reported that the younger children displayed varying degrees

of more complex social behaviour describing some younger children as showing

"greatly elevated levels" of such behaviour (Lougee et al., 1977, p. 1360).

Goldman's (1981) two-part study focused on: first the amount of time that

children aged three and four years old, in same-age or mixed-age groups, spent on social

participation; and secondly the age relationships in those groupings. Data were

collected by conducting ten, three-minute observations during the children's free-play

time. Results of the first part of the study revealed that in mixed-age groups the three-

year-olds (the younger children) spent less time engaged in parallel play than those

three-year-olds who remained in the same-age group. Parallel play is, in terms of social

content, a less complex form of play than associative or cooperative play (Parten, 1932).

It can be sustained with little or no peer social interaction. Conversely, peer social

interaction forms the basis of associative or cooperative play. The results of Goldman's

(1981) study support the current hypothesis that the presence and activity of an older

peer are likely to facilitate the younger child's engagement in more complex play.

Outcomes of the second part of the study suggested that, in all groups, the sex of

the older peer rather than age was the more influential factor in the choice of a play

partner. The third condition in the present project is dyad play in which the toddler

participant is paired with a 'self-chosen' familiar preschool-aged peer. Goldman's

(1981) research findings suggest that this older, familiar peer is likely to be the same sex

as the toddler participant.

32


Toddler interaction was the focus of an investigation by Rothstein-Fisch and

Howes (1988) whose research was conducted in Family Day Care homes. In such a

setting children can make the choice of interacting with others who are either younger

or older than they. Rothstein-Fisch and Howes (1988) examined partner preference and

the complexity and content of play. Six observations, each of five minutes were

undertaken and six social behaviours including initiation, imitation and the offering,

receiving or taking of toys, were noted. The toddler participants' (aged from 16 to 23

months) preferred play partner was 25 to 35 months old. The play engaged in with

these two-year-old partners was more complex peer play than that shared with other

partners. Preschool-aged peers were watched more than others by the younger

participants but they most often imitated their two-year-old companions.

Toddlers' peer social skills, including maintenance of interactions and

complexity of initiations, were investigated by Brownwell (1990). The question posed

was whether young children aged 18 and 24 months could modify or adjust their social

behaviour to that of a differently-aged partner. Dyads comprised of 31 children of each

age (18 or 24 months) who had been randomly allocated to a familiar same-age or

older-age peer. Almost all the dyads were same-sex. Observations were undertaken in

a room set apart from the children's playroom or classroom. A range of toys and play

materials chosen to elicit a variety of play and interactional responses was provided.

Once the participants appeared settled and at ease, 15 minutes of videotaped data were

collected.

Brownwell's results are varied and interesting. For example, older more sociable

peers whose behaviour Brownwell described as "affectively positive" (1990, p. 845)

appeared to elicit more interactive and positive behaviour from the younger, 18-month-

old toddlers than was shown by those toddlers toward their age-mates. The 18-month-

olds also employed more sophisticated initiation behaviour toward the two-year-old

children than that shown toward their age-mates. A second finding was that the older

participants, toddlers aged 24 months, utilised lengthier and more complex initiations to

younger, 18-month-old partners than to their age-mates. A third important finding was

that the younger toddlers imitated their older, 24-month-old playmates more than they

did their same-age partners. Imitation of the older peers' play activity in two

conditions, mixed-age play and dyad play, is considered in this thesis.

Clearly Brownwell's findings indicate that very young children can adjust their

interaction patterns to accommodate those of older peers. The younger toddlers'

33


responses suggest that partnership with an older peer is also stimulating and motivating

for the younger child (1990). A n important variable in a study of this kind may be the

temperament or attachment security of the older child. As cited earlier in this

discussion of Brownwell's (1990) study, positive outcomes were reported for those

toddlers w h o engaged with older children of a more sociable nature who also exhibited

more 'affectively positive' behaviours. Sociability and positive affect are 'within-child'

characteristics likely to facilitate and support positive peer interactions (Thomas, Chess,

& Birch, 1968).

Bailey, McWilliam, Ware, and Burchinal (1993) continued the theme of

investigating the influence of same-age and mixed-age grouping with a project focusing

on children's social interactions with peers. Their research method involved observing

eight groups of children, four of same-age peers and the other four mixed-age groups.

Participants ranged in age from 18 to 60 months. As in the earlier study by Blasco et al.

(1993) - reviewed in part 2.2 of this chapter - each group included two children with

disabilities. Four videotaped observations were made each lasting 15 minutes. Analysis

showed that the younger children spent approximately 67 percent of their time

interacting with older children. In contrast the older children spent almost 75 percent of

their available time interacting with their same-age peers. Bailey et al. (1993) surmised

that the latter outcome might have resulted from the long-term familiarity that already

existed among the older children, as the child-care population from which the groups

were drawn was very stable. Such familiarity may relate to the current study since the

partner in the dyad play condition is the participant's 'self-chosen' familiar older peer.

2.4 LITERATURE EXAMINING YOUNG CHILDREN'S IMITATIVE

BEHAVIOUR

Observational learning or imitation of observed behaviours was recognised by

Bandura (1977) as a category of social learning. H e suggested that children were more

likely to imitate a familiar and well-liked model. A s Bandura (1986) modified his

theory by incorporating the notion of cognition as a mediator in the 'learning-through-

imitation' process, so he postulated that children become more selective than he earlier

thought when choosing a model.

As mentioned earlier, imitation is thought to be a key component in the

development of symbolic play - from a Piagetian perspective imitation is a feature of

34


practice play which is a forerunner to emerging symbolic play skills (Piaget, 1962).

Support for the consideration of imitative behaviour is strengthened by the more recent

work of Tomasello (1995). H e describes imitation as a "relevant phenomenon" in

relation to joint attention (1995, p. 110) in that it constitutes a "following into adult

behaviour and attention" (1995, p. 110).

The findings of each of the studies presented in Table 2.3 (and described briefly

following that Table) offer useful supplementary information and guidance in support of

this prediction.

35

CO

00 3 XI 3

CO

y\

CO

ea 4)

s 4)

"S > * - p

3 4) X) 3 4)

a 4) XI 3

CO

S5-§

4) r"

fc < E

60 3

'3 4) CO

Author

(s)

Date

More linguistic and

gestural imitation by children

in home-care.

,—__

—-

4 linguistic stimuli

4 gestural stimuli

Linguistic and gestural

stimuli modelled -

home-care and

day-care

settings

•UP

I -̂* UP

4)

g" <u

48

18-48

LDCCC

Home

Francis &

Self

(1982)

Prompted

pretence in

non-pretend players and

those who had few

acts of

pretence

in pre-modelling

assessment.

Action &

linguistic expressions

of pretence (adapted from Fenson,

1984)

oo

.£ 13 Xt

o E 0-P

O CO 4-P

o 4) fc

w 3

00 y-(

X p — 1

33

30-42

LDCCC (but

aside from the

regular

program)

Lyytinen

(1989)

Children and enculturated

chimpanzees showed more

imitation behaviour than 'mother-

raised' chimpanzees

Imitative learning - 'simple' and

'complex' actions

Deferred

imitation

Modelled

novel actions

of events

1 experiment

(5 components)

16

18-30

Laboratory

playroom

Tomasello,

Savage-Rumbaugh,

Kruger

(1993)

Participants able to engage in

deferred imitation after 48 hours

and in a different context.

3 O 4-P

ea 4-1

X) 4)

<<-p

4)

Q

Peer

imitation of

modelled

actions

3 experiments

involving

imitation of an

'expert' peer and a

change

of context

Laboratory

playroom

LDCCC

Hanna &

Meltzoff

(1993)

CO

00 3 XI 3 'Uy

CA

I CO

ca

3 4J

'S >

1 1

CA

55 •§

4) _c OB'S

fc < fi

00 3

3 4) 00

Author (s)

Date

Positive relationship found

between joint attention and

imitative learning.

Visual attention & behaviour

towards object of focus

Joint Attention Episodes (JAEs)

Modelled

novel actions

of events

1 experiment

(5 components)

oo \D —i

Laboratory

playroom

Carpenter,

Tomasello, &

Savage-Rumbaugh

(1995)

Recall

skills and organisation of

recall similar to older children.

Recall of causal events, sequences

in causal events and

arbitrary

events

Modelled, novel

sequence

of events

CO

8

1 u 4> 4) CN

oo y yy P—H

Laboratory

playroom

Mandler &

McDonough

(1995)


While the toddlers' likely imitation of a more-skilled older peer is the focus of

this part of the literature review, their imitation of activities and behaviours modelled by

adults also warrants discussion. Staff, especially those trained in early-childhood

education, influence the development of children's symbolic play in a number of ways

including: the toys, materials and equipment they supply to support children's

developing symbolic play; the attitudes they themselves display toward symbolic play;

and finally their o w n engagement in symbolic play with children. The last offers

children the opportunity to engage in imitation in play that is likely to be more complex

than they would otherwise exhibit.

Francis and Self (1982) investigated the imitative responsiveness of two-year-

olds in L D C C C s and home care - particularly examining the importance of the

caregiver-to-child ratio. Each participant was seen alone and the researcher modelled

(for approximately two seconds only) eight stimuli - four of which were verbal and four

gestural. A ten-second period was then allowed during which time the toddler had the

opportunity to imitate the proffered stimuli. N o encouragements were given either to

elicit or commend a response from the participant. Responses were scored in the

following ways: as a perfect imitation of the stimuli; a partial imitation; or no response.

Results of the research conducted by Francis and Self (1982) confirmed the

findings of earlier research in that children in home-care displayed significantly more

imitative acts - the effect of the setting was significant. The results themselves are,

perhaps, not surprising but implications for the current study arise from these findings.

The process by which young children's imitation develops is thought to start in infancy

when initially the caregiver imitates the baby's sounds and gestures (Berk, 1997, p.

140). In time the infant's imitative responses are generalised to others in the

environment. In a L D C C context, where the child-to-caregiver ratio is higher than that

in home-care, it m a y be that others available for imitation, for example older peers, play

an important role in the young child's acquisition of skills.

Implications arising from Francis and Self (1982) study relating to child-to-

caregiver ratios in L D C C C s are explored in part 2.11. of this literature review when a

range of components (related to the quality of the provision of care) in the L D C C C

environment is examined.

Lyytinen's (1989) research investigated the effects of modelling on the actions

and verbal expression of the symbolic-play behaviour of 33 children aged between 30

and 42 months. A symbolic play pre-test assessed the children's current symbolic-play

38


status enabling the participants to be identified as being at one of three levels of

symbolic play, the lowest of which was the 'non-symbolic' players. Decentred' players,

those children who engaged in decentred activity, were classed as the most sophisticated

symbolic players with 'transitional' players, those with the ability to engage in object

transformation (the use of an object to stand for or represent another), placed in the

centre, between 'non-symbolic' and 'decentred' players.

A n 18-minute videotaped observation was made comprising first of a seven-

minute period when participants played freely with a range of replica toys. Next, the

adult modelled three symbolic-play sequences - a doll was used as a passive participant

in the first two modelled segments but was depicted as an active participant (or agent) in

the final sequence. (The adult combined 'child-appropriate' language with the

symbolic-play actions.) The adult demonstrations of symbolic play lasted

approximately four minutes. Following the adult's demonstration, the children were

videotaped as they continued playing for a seven-minute period or longer if they

wished. (Therefore the total length of the video-taped observations was at least 18

minutes.)

Interestingly, in the children's play that followed the adult-modelled sequences,

non-symbolic players' 'self-as-agent' symbolic play decreased and symbolic play

involving action upon other objects increased significantly. Lyytinen (1989) found that

the 'transitional' symbolic players (those who were able to engage in 'transformations')

made the most gains while the 'decentred' players' activities remained unchanged.

Moreover, a significant increase was seen in all children's language, across all groups,

following the modelled behaviours.

The modelling or demonstration of novel events or actions, which may be of an

arbitrary nature or, if involving a sequence, suggest a causal order, have been

components of studies by a number of research associates (Tomasello, Savage-

Rumbaugh, & Kruger, 1993; Carpenter, Tomasello, & Savage-Rumbaugh, 1995;

Mandler & McDonough, 1995).

The first study in this group to be reviewed is that of Tomasello et al. published

in 1993. They investigated the role of imitation in the learning of modelled novel

actions on objects, by both young children and "enculturated" (Tomasello et al., 1993,

p. 1688) - that is, raised in a 'human-like' cultural environment, chimpanzees. It was

hypothesised that the enculturated chimpanzees would display imitative learning while

those raised by their mothers - lacking the social environment experienced by the

39


enculturated chimpanzees - would not display imitative learning behaviours (Tomasello

et al.).

Eight of the human participants were aged 18 months and eight were 30 months

old. Six chimpanzees, three of which were raised by their mothers and three of which

were enculturated, took part in the study. First, the experimenter offered four, of the 12

objects to be presented, to the participant for free-play activity lasting approximately

four minutes. Strategies were used by the experimenter to encourage play and

exploration if the participant appeared reluctant to explore or play with the objects. The

first object was then used, by the experimenter, in a simple action and then offered to

the participant with the instruction for them to repeat the same action. T w o

demonstrations of the simple action were made if necessary followed by a

demonstration of the correct or conventional use of the object.

The second stage of the procedure required the experimenter to present four new

objects to the participant for free play - again for approximately four minutes. The

correct use of each of the objects was then modelled by the experimenter and the object

removed immediately after this demonstration. T w o days later the third and fourth

components of the procedure were carried out replicating the first two parts of the

procedure. O n that same day the final component of the procedure, involving the

presentation of the objects from the second part of the process, was implemented. This

involved participants' being presented with the same objects as used in part two of the

procedure. Objects were presented to participants one by one following the same order

as in the earlier presentation. The behaviour of the participants with each of these

objects was observed for two to three minutes. N o guidance, either verbal or non

verbal, was given at this stage.

For the current project, the findings of Tomasello et al. (1993) related to the

human participants are of most interest. In summary the 30-month-old children and the

enculturated chimpanzees used the objects in complex actions during the free-play

period (that is, before the action had been demonstrated) more than the younger children

or the mother-raised chimpanzees. Children aged 30 months are very skilled at playing

appropriately with realistic props representing objects observed in their environment.

A n increased capacity in both attention and cognition allows them not only to remember

their experiences but also to note and remember the components of those experiences in

some detail (Westby, 1991). However, a second result pertinent to the present research

is the engagement of the enculturated chimpanzees in that same appropriate behaviour

40


(Tomasello et al., 1993). This finding underscores the important role played by the

sociocultural environment and the components therein - objects and artifacts - central

to the theory of Vygotsky (1978) and more lately of Rogoff (1990). Each of these

theorists stresses the powerful influence exerted by both the social context and those

more-skilled others within that context in informing and supporting younger, less-

skilled children's learning and development.

A subsequent study undertaken by Carpenter, Tomasello, and Savage-

Rumbaugh (1995) serves to further emphasise the importance of the social environment

as revealed in the study by Tomasello et al. (1993). A portion of the video-taped data

from the study by Tomasello et al. was recoded focusing on the participants' 'looking

patterns' and their gestural behaviour as they interacted with the experimenter.

Participants in this case were the three enculturated chimpanzees, the three mother-

raised chimpanzees and six of the 18-month-old children. A second goal for Carpenter

et al. (1995) was to identify whether, for any group of participants, a relationship

existed between joint attention and imitative learning.

Tomasello defines joint attention as occurring when "the child coordinates her

attention to the object and adult at the same time that the adult coordinates her attention

to the same object and the child" (1995, p. 107). The results of the research undertaken

by Carpenter et al. (1995) were significant in that the imitative behaviour for all

participants was positively correlated with the percentage of time spent in joint attention

with their more skilled partner.

Support for the methodology of the present project, in which a toddler

participant was paired with an older, familiar preschool-aged peer for dyad play, is

found in these results (Carpenter et al., 1995). Materials to elicit symbolic play were

provided for the dyadic condition because it was predicted that the younger children

would imitate the older peers' more sophisticated symbolic play. Joint attention

facilitates that imitation process.

Further support for considering the role of imitation as likely to be central in the

participants' symbolic play activity in the current project is found in the work of Hanna

and Meltzoff (1993). Their study involving three experiments examining the 14- to 18-

month-old participants' deferred imitation of activities demonstrated by a 14-month-old

'expert' same-age peer. The final trial showed that the toddler participants were able to

reproduce the modelled events first, after a 48-hour delay and, secondly (although the

41


demonstrations had occurred in the laboratory playroom) in the context of their

L D C C C .

The findings of a study examining infants' observational learning and reported

by Mandler and McDonough in 1995 are also pertinent to the current investigation.

These researchers investigated the ability of 11-month-olds to recall novel modelled

sequenced events on two occasions after the demonstration period.

T w o experiments were conducted by Mandler and McDonough (1995): the first

entailed appraising the infants' recall of the demonstrated sequences after just 24 hours;

and the second required the same appraisal after three months. Four two-step events

were modelled: two were causally ordered in that after an 'enabling' event an 'end-state'

occurred; and two were arbitrarily ordered (1995, p. 460). N o verbal support was given

during the modelling session or recall session by either the infant's parent (always

present) or the researcher.

W h e n the demonstrations were completed in the first experiment, materials from

one each of the causally ordered and the arbitrarily ordered events were handed to the

participants after a delay of 20 seconds. A n assessment of the infant's activity with the

props was made. Twenty-four hours later, the materials that were used in the

demonstration were presented to the infants, one after the other, and again their activity

with the materials was assessed. Assessment of recall was gauged by observing the

participants' actions (when the objects were re-presented) and the sequence in which

these actions were conducted. After the 20-second delay, assessment indicated infants

could recall approximately half the modelled actions - both causal and non-causal.

Results of this component of the study indicated that the participants effectively recalled

two-step sequences - the causal events - but not the arbitrarily constructed sequences,

after a 24-hour delay.

In the second experiment a baseline was established for each participant as he or

she encountered the materials again. A control group was introduced at this stage and

the same procedure as used for the first experiment was then followed. The original

participants were then assessed, as in the first experiment, after a 20-second period and

again after 24 hours. They showed a significantly higher rate of recall on causal events

than the control group both when the baseline was established and after the

demonstrations. The arbitrary events were not recalled by the original infants before

modelling, nor after modelling.

42


The participants in the Mandler and McDonough (1995) study were 11 months

old at the time of the first experiment and 14 months old when the second experiment

occurred. At the time of Mandler and McDonough's (1995) first experiment, the infants

were in their pre-verbal stage of language development. However, at fourteen months

of age (when the second experiment was conducted) some language, though likely to be

single word usage, was evident. The results of Mandler and McDonough's (1995) work

illustrate the notion of deferred imitation as a manifestation of non-verbal recall.

The major hypothesis for the current study is that toddlers who are regularly

engaged in free play with older, preschool-aged peers (whether group mixed-age play or

dyad play) will exhibit symbolic play more frequently and at a higher level than when

they are engaged with same-age peers. It would seem in considering Mandler and

McDonough's (1995) research that the toddler participants involved in the current

project are likely to bring with them skills for acquiring and maintaining the enhanced

play behaviours that they are likely to witness when engaged with more-skilled players.

In accordance with Mandler and McDonough's results it appears likely that the

enhanced symbolic-play behaviours will be observed in Dimension T w o ('thematic

content') and Dimension Three ('organisation of themes').

2.5 PROGRAMMING CONSIDERATIONS

The research examined so far has tended to favour the grouping of younger

children with older peers in multi-age or mixed-age grouping. There exists, however,

evidence that there might be an optimal range of age and stages of development

successfully catered for in mixed-age grouping.

Sundell (1994) reviewed the findings of several studies he undertook, between

the years 1989 and 1994, in Sweden, a country that implemented mixed-age grouping

throughout its L D C C C s and primary schools more than twenty years ago. Mixed-age

groups in Sweden tend to span larger age ranges than in other countries and most often

include children aged from one year to seven years. (Children start primary school later

in Scandinavia so may be at preschool until seven years of age.)

Sundell's (1994) collective findings point to the need to recognise an optimal age

span for successful mixed-age grouping in contrast to the span of up to seven years.

The latter he concluded is inappropriate for a number of reasons. Sundell suggested a

number of disadvantages might arise from mixed-age grouping for both children and

43


staff - for example, the needs of the younger children disrupting the older children's

activities. From the teachers' perspectives, mixed-age groups also required more time

for planning and placed a greater intellectual demand on them. It appears these

disadvantages may stem from the excessively wide span in children's ages.

Sundell concluded that a reduced age span for mixed-age grouping might be

warranted, though no suggestions regarding what age range could be were put forward.

Interestingly, in Sweden, a reduction of 30 percent in the number of mixed-age classes

occurred in the late 1980s. Simultaneously, the practice of all children spending some

part of the day in same-age groups was introduced (Sundell, 1994).

A second implication arising from Sundell's work concerns the role played by

imitation in the development of symbolic play. The age-range of children attending

L D C C C s in Australia is from approximately six months to 66 months. Children

younger than 18 months were rarely included in mixed-age groups in the participating

L D C C C s ; consequently the age-range available (approximately 18 months to 66

months) provides an ideal opportunity to examine imitation in a more restricted range of

ages.

Sundell (1994) also referred to research conducted in conjunction with a co-

researcher, Johansson, (1989, but published in Swedish only) involving the responses to

a questionnaire of almost 900 teachers in more than 400 early-childhood settings. The

questionnaire elicited information concerning the age-span of the sibling-groups, the

teachers' instructional styles and the children's development. Johansson and Sundell

(1989; cited in Sundell, 1994) concluded from the teachers' responses that in sibling-

groups, where the age-span between the youngest and the eldest child was large,

children's development was 'poorer.' This finding seemed pertinent for both the

younger and older peers in the mixed-age groups in that the younger peers appeared not

to be well supported by the older children in the group while they, the older children,

were under-stimulated. Clearly, successful and appropriate programming for a mixed-

age group must cater for all developmental levels represented within that group.

Such considerations reflect the content of yet another question raised in the

present study - that of programming. It is likely that programming is another

component of the L D C C C environment that influences the development of toddlers'

symbolic play.

In summary, when reviewing the literature cited in this section, the efficacy of

grouping toddlers with older peers, as a way of enhancing toddlers' learning and

44


development, would seem likely to be a successful strategy to employ in supporting the

younger children's development. The findings described in the studies reviewed

illustrate the younger child's potential for acquiring various social, cognitive and

language skills as a result of interaction with the older child.

2.6 LITERATURE EXAMINING SOME ADULT INFLUENCES ON

TODDLERS' SYMBOLIC PLAY

The adults who feature in this research are mostly and perhaps not surprisingly

parents. Most often it has been the mother's interactions that, with respect to the

development of toddlers' pretend play, have long held an interest for researchers.

The term caregiver when used in the current project refers to the non-familial

caregiver or teacher present in the L D C C C . It is likely, while taking into account the

differing contexts, that the manner in which such caregivers support the participating

toddlers' symbolic play development in L D C C C s is likely to be similar to that of

mothers. In the case of caregivers trained in Early-Childhood Education, it is expected

that a more 'informed' practice of supporting and responding to young children's

symbolic play would be evident.

The questionnaire distributed to the toddlers' parents required them to first

describe briefly the developmental expectations they had for their child and secondly a

typical 'week-end' experience involving their child with others (parents, sibling, friends

or others).

Therefore this section of the literature review contains two parts - the first

appraises a range of studies focusing on mothers, and in two cases fathers, as an

influence on children's symbolic-play development. The second part contains a small

range of research studies, which consider the influence of the L D C C teacher or

caregiver's interactions and teaching style on young children's learning and

development. The importance of teacher training in Early-Childhood Education is also

considered.

The mid-1970s witnessed a burgeoning of research interest in the interactions

between parental caregivers and their children and the effect or influence of these

interactions in supporting the very young child's development. One of the stimuli for

such interest m a y well have stemmed from Bruner's (1974) notion that developmental

benefits were accrued by young children engaged in play with a 'more-skilled other'.

45


He believed that the optimal condition for learning was a playful context - this being

synonymous with a low-stress environment. The maternal parent was both as accessible

and as available as the more skilled other, able to employ such strategies as scaffolding

or tutoring in her interactions with young children. Researchers were keen to explore

these intimate mother-child interactions in search of implications for a range of

developmental areas including young children's symbolic thought and activity.

The studies presented in this component of the literature review have been

selected as a representative range of the projects investigating the parental adults' role

in influencing young children's symbolic-play behaviour. The research reviewed spans

the twenty years from 1977 to the present. Methodology comprised two main

approaches: namely the child playing alone and then playing in the company of an adult

(most often the maternal parent) either in the home under naturalistic conditions or in

laboratory play-room conditions. While other types of play, for example exploratory or

combinatorial play, may also have been investigated by the researchers, for the purpose

of this discussion symbolic play remains the focus.

In the current study the second related research question considers the symbolic

play of first-born toddlers compared to that of toddlers with older siblings; therefore,

studies of outcomes for first-borns are especially pertinent to this review. Included in

this group is the work of Damast, Tamis-LeMonda, and Bornstein (1996), Bornstein,

Haynes, O'Reilly, and Painter (1996) and Haight, Parke, and Black (1997) - all of

w h o m conducted their research with participants w h o were first-borns.

These studies are listed in Table 2.4. A brief summary of conclusions and

particular issues bearing on the present research follows the Table.

46

CO

60 3 • *-( •d 3

CO

1 ea 4)

s 4)

'S >

1 1 •a HH

1/1

25 "§

Z < fi

60 3

s 1/5

Author(s)

Date

Symbolic

play sequences

initiated by

child completed if 'joint attention' with

mother

occurred.

(If child alone sequences

rarely

completed.)

Actions with objects (Inhelder et

al., 1972)

Symbolic

activity (Authors'

scheme)

Social context - mother

available

B o Xi X <N

24

18-24

I yy

Dunn&

Wooding

(1977)

Diversity of play increased with

mother's

presence.

An increase in symbolic play at 28

months of age when mother present.

Play categories:

exploratory, combinatorial,

symbolic and ambiguous

(Authors' scheme)

Social context - alone at

home vs with mother in

labaratory

1 x 15 min.

in the home

1x5 min.

in lab.

o o en CN

Home

and laboratory

playroom.

O'Connell

Bretherton

(1984)

Pretend play with materials of both

high and low representational details

increased if adult play suggestions were

made.

—

Checklist of 5 play categories,

each category had two classes

(Authors' scheme)

Social context - play

with high and low

representational materials

with and without adult

play suggestions

-3-

20

20-33

Not defined -

description

suggests

laboratory

Playroom.

Steele &

Hrncir

(1985).

The level and duration of symbolic play

increased when the mother was

involved

in the play.

Symbolic

play (McCune-

Nicolich, 1977)

Maternal involvement (adapted

from McCune, 1984)

Social context - child

playing alone (20 mins)

and with mother (10

mins)

7 x 30 min.

(4 of which

used

in these

findings)

VO

Laboratory

playroom

Slade

(1987a)

60 3

1

3 ea

3

= & 4) ea °"

fi 8'2 rt 13 s » S S

B S o UP e3 U

d g *

IP cS 3 O

- 6 0 - g

00 00 O N

ea 4-1

4> 'ul 4) H

E o

vo

r s a £ •*

"ea 3

13 O

g •» £ ea

o € .B 53 eo co

a g

.§ i i e • ^

4-P T j

3 3 5 e3 o «C

'5 'S « 2P OH «

o ea

3 O

g CO

p.

XI 4) j>

"3 > 3 -

co .3 u co 4-» ** 4-P

O CO 4> 60

2 § g 3 3 .3 Op.tJ

2 ^ J3 >>

> s

4> ..—,

o .O X o » ea 60 O B -xi

^ 2 S J ex £

CO

O

E X XJ 4) B 0. 0l

y\ 2

8 a O c 2 XI fi 4) co — ea O

I-st u

-H..S .2 o o fi

1 g CO 5

13

> S3

o 3

ii %3 •g'l B -* JB 'a > 60

x B M S "o. "u

8 I

3 4)

G a o 13 > 4) X) ea O .O

o a. D. 3 co >^ 13 3

"4—p

o

4)

S3 CO

ea 4)

s

B r-~

.2 o\ 4-1 i-H

ea . O co •a u CO i£

S o ea B

"Eg CUP CO

O 3

" -2 ea 3 CO c_t

T3 fi 4) p

co en S-.00 rt Ov

& li E: 4) 3 Ov c ea _<

t=; -0 s ° 13 yy ea « co 3 Xj o

4) J3 X3 co .13 -UP

'fi 3 >,

ca a. 4-p

o CO

3 O CO

3 4) fi TJ

ca o o

11 13 3 41

u •n 3 UP

O (1) > ca 4)

a n o CO o

•a E O 4) ea J3

U CUP co ° 3 co > O M

60 "u

<2 o ta rg « 3 "* < CN w

§4 cfc § •rt eo

4) 3

>—-1*-co Ov 4) .—i • yP

O "rt 60 ^

ta B o 5 & 3 0-. HH

=y

J3 o

O o

. p — I

° yy U T3 ca c

6" ov oo r~ ov ov >M3 e^ ? C^

fi £ Is

4)

I 'S > 4-1

3 CJ

•a

g

CO

13 ea

cd .S * >v-3 ts •a r2 o -a ' co u S 4) M 43 60 3 o ea fa co i ' y* ^ 3

Sill Py <y £ -S

iS X y\

4) ̂ 3

° 60

o J2 co a

60 3

S '5 — CO

^ 5 > 3 > O

B°fi

I00 13 T3 r3 4) X! 60 cj ea

T3 3 60 ea 3 a -S 3 eo

B « 13 4) I .2 5 "O '-S >< 60 ' E -a B e 53 u, o

8P11^U3

"3 2 y- X- S O J3 "g -g ?. CO CJ > ? co

3 4) 3 O

a o CJ ca y-y*

O

CO

in

2 o .—1

x tN

3

fi

O yy

xs 13

(N S

VO

X

ea CJ

«n 'o 4)

a CO 4-P

O 3

4) n 60 Xi < fi

OV VO

in Ul

p — ( . ^ o CN

00 CN CN CN

•t CN

IT) y

•s >

3 4) XI B 4) CX U XI

B IE 4—i

ca

x

60 3 •Z3 4-P

4) CO

4)

fi 4)

fi o

ea o

o

IS „ T J « 00

u o S 2 H P P O C *

73 ov

Q C

co OV •O Ov ' * ^ y-^

Uy C

o B 4) XI *

CO

60 3 T3 3 'CU

CO

CO

ea 4)

4>

"5 > 4-P

g X

g

£ c HH

s-8

£ < e

60 3

'fl 4) CO

Author(s)

Date

Mothers directed pretend play toward

12 month olds.

By 24 months children initiated about

50% of pretend play sequences toward

mother.

Pretend play

Social conduct

of pretend play

(Authors' scheme)

Social context - with

mother

7x3-4

hours

(at 12,

16,20,

24, 30,

36 and

48 months)

9

12 at

start

48 at

end

1 o

Haight &

Miller

(1992)

Pretend play initiated by mothers

engaged in 'everyday'

activities both

caregiving and household.

Mothers used pretend play to 'manage'

child's behaviour.

"Children initiated pretend play ... for

'fun'... and from contexts other than

play" (Haight

et al. 1994, p. 520).

Activity context codes (Authors'

scale)

Social functions (Authors'

scale)

Social functions of

mother-child pretend play

3-4 hours on

4 occasions

9

12 at

start, 48

at end

I o

Haight,

Masiello,

Dickson,

Huckeby,

& Black

(1994)

Symbolic

play increased with child's

age and with father's involvement.

Pretend play themes with father were

gender-typed.

Play with objects (adapted from

O'Connell et al. 1984)

Social play (adapted from Peer

Play Scale, Howes, 1980)

Cooperative pretend play

(adapted from Howes, 1985)

Play themes (Author's scheme)


playing alone and with

father

1 x 40 min

(20 mins

alone,

20 mins

with father)

60

(30

Anglo-

American

30

Mexican)

16x18

16x24

Laboratory

playroom

Farver &

Wimbarti

(1995)

CO

60 B X 3 Uy

CO

—̂'

I CO

ea 4)

s -2 ca

S3 >

1 X

g & X 3

yy

CO

S3 Z < 6

00 3 'yt 4-4 4-1

4> CO

Author(s)

Date

Mothers adjust their play to either the

same

level as the child's or higher.

Exploration

Non-symbolic

play

Symbolic

play

(Authors' scheme)


playing with representat

ional toys and mother

o X

O —i •O CN

4)

E o X

Damast,

Tamis-

LeMonda

Bornstein

(1996)

Children's symbolic play increased

when mother was

involved in the play.

Eight levels of play, plus 'no

play' (Authors' scheme)


playing alone and

with

mother

2x10 min

(1 x 10 min

alone,

1x10 min

with mother)

5 o —< CN

4)

fi

Bornstein

Haynes,

O'Reilly,

& Painter

(1996)

Pretend play comprised 11%-22% play

content of each observed play session.

Parents and

children both respond

to

and elaborate on

each other's

initiations

of pretend play.

Pretend play

Rough and tumble play

Book reading

(Authors' schemes)


playing with mother

and father (separately)

210 min

(3 x 35 min

with

mothers,

3 x 35 min

with

fathers)

29

24, 30, 36

8 o X

Haight,

Parke, &

Black

(1997)


While each of the studies presented above in Table 2.4. has some relevance to

the current research project, a closer inspection of three particular studies is useful - the

work of O'Connell and Bretherton (1984), Slade (1987a) and Haight et al. (1994).

O'Connell and Bretherton (1984) observed 30 participants involved in a 15-

minute independent play session at home and a five-minute collaborative play session

with their mothers in a laboratory situation. Each participant was observed twice in

each condition - first when aged 20 months and again when they were 28 months old.

The following four play types were examined: ambiguous, symbolic, exploratory and

combinatorial play. Differences were found to exist in the play manifested in the two

play sessions and these differences were attributed to the mothers' influence when

engaged as a play partner (O'Connell & Bretherton, 1984). This influence was judged

to be significant in all four play types, but the diversity of exploratory and combinatorial

play, in the collaborative sessions, was greater than that observed in the two other play

categories (ambiguous and symbolic).

In keeping with later research - for example, Farver and Wimbarti (1995) -

while changes in the children's play behaviour would seem to be a result of the adults'

input as play partners, play behaviour also remained a function of the children's

developmental stage. This was illustrated by increases in diversity in exploratory and

combinatorial play for the younger children aged 20 months and in symbolic play for

the 28-month-olds (O'Connell & Bretherton, 1984).

T w o of O'Connell and Bretherton's observations are of particular relevance in

terms of the interpretation of Vygotskian theory as espoused by Rogoff, Mistry, Goncu,

and Mosier (1993). The first is that the researchers proposed that the child, by heeding

some and ignoring others, actively determines the effectiveness of the mother's

interactions. Such an observation suggests that the quality and content of the play

arising in collaborative sessions depends to some degree upon the sensitivity of the

adult play partner to the play situation. Interactions such as these would fall within the

Vygotskian framework of the 'zone of proximal development' in that the more-

experienced other (in this case the mother) is able, as co-constructionist, to support the

child and facilitate the development of greater competency or skills.

The second observation is that the researchers suggest that children were also

discriminating in that they accepted hints or directions for which they appeared to be

'ready'. This feature, together with O'Connell and Bretherton's (1984) conclusion that

the play behaviour exhibited by the children remained a function of their developmental

51


level, supports Vygotsky's (1962) claim that maturation processes limit potential for

progression.

Burner's (1972) concept of the transmission of cultural rules and conventions

through play, though grounded in ethological theory, is not dissimilar to that of

Vygotsky. H e postulates that a 'tutor proneness' - which may be likened to a

predisposition to learn from adults - exists in very young children. Yet another

proposition by O'Connell and Bretherton (1984) congruent with both Vygotskian theory

and an ethological framework is their view that the child chose to act upon the adult's

suggestion when the activity was in keeping with current tasks the child was trying to

master. Conversely those propositions offered by the mother, that reflect activity

already mastered or well beyond the child's present capacity, are dismissed or not acted

upon. The conclusions of this study by O'Connell and Bretherton (1984) have

implications too for a Piagetian theoretical perspective in that the child is actively

involved in deciding what elements on offer in the environment are useful to continue

the process of knowledge construction through assimilation and accommodation.

A longitudinal study conducted over a period of 12 months by Slade (1987a) not

only supported the work undertaken by O'Connell and Bretherton (1984) but added to

the diversity of findings. Slade observed, through the use of film, the interactions of 15

mother-child pairs over a period of a year - starting when the child was 18 months old

and concluding when the toddler was 30 months of age. The dyads were studied every

two months, the focus being symbolic play with each half-hour play episode coded as

either 'child-initiated' or 'mother-initiated'. Also coded were the length of the episode,

the level of play displayed and the degree of maternal involvement.

Results indicated that the level of symbolic play increased and was more

sophisticated when the mother was available and engaged in the play. However, the

complexity of the symbolic play exhibited was in keeping with the child's

developmental level. Interestingly, the length of the symbolic-play episode remained

unchanged until the child was 28 months old even with maternal involvement; and

again the added duration of episode was considered to be more a function of the child's

age and developmental stage than a result of maternal involvement (Slade, 1987a). In a

further analysis of the same data set as used in the study reported on here, Slade (1987b)

found that securely attached participants scored significantly higher on symbolic-play

measures than those participants not securely attached.

52


Haight et al. investigated "how pretend play is embedded within everyday

routines, and the social functions of its initiation" (1994, p. 519-520). Observations

were made (in the home) when the participants were 12, 24, 36 and 48 months old and

subsequent analysis showed that mothers initiated bouts of pretend play at times other

than those dedicated to 'just playing'. They incorporated pretend play into everyday

care-giving routines and tasks and interestingly about 33 percent of the mothers'

initiations of symbolic play were to manage children's behaviour - perhaps somewhat

instrumental in nature. Importantly the participants also initiated bouts of symbolic play

in routine task situations.

In summary, the findings of each of the research programs cited in Table 2.4

indicate that, when children are in the company of a primary caregiver, pretend play is

likely to occur. The complexity, diversity and frequency of symbolic activity are

influenced by the nature of the adult's collaboration and the child's developmental

stage. Other factors enhancing symbolic play include both the adult's and the child's

elaboration on and response to each other's initiations of pretend-play activity and the

adult's modelling of or suggestions for tasks involving pretence. The caregivers

engaged in pretend play with children, supported that play by adjusting their level of

symbolic play to accommodate that of the children.

Guidance for the current research is sought from these findings because the

participants in this study spend a minimum of 24 hours a week in L D C C C s . Each of the

participating centres maintained the practice of 'primary' caregivers - caregivers

regularly responsible for the care of the same small group (when possible) in a bid to

maintain stability and promote close responsive relationships between the carer and

child. It is suggested that such a relationship, especially when the carer is trained in

Early-Childhood Education, may in part resemble those of the mother-child dyad in

supporting developmental outcomes for the toddler participants.

2.7 LITERATURE EXAMINING THE INFLUENCE OF OLDER SIBLINGS ON

THEIR YOUNGER SIBLINGS' PLAY

The second category of literature to be examined includes those studies arising

from the investigation of the influence of older siblings on their younger siblings'

development of play skills and interaction skills. A n interest in reviewing such studies

arises as one of the additional research questions posed in the present project is whether

53


toddlers with older siblings exhibit symbolic-play skills more frequently or at a higher

level than those toddlers w h o do not have older siblings.

Before widespread industrialisation, it seems likely that children's acquisition of

new skills and learning largely took place in the context of the family, a more extended

entity than in developed countries now, and the community. Families and

neighbourhood communities provided an environment which contained a number of

possible teachers or tutors (each of w h o m were likely to be at different stages of

development) older siblings were among those tutors available.

Cicirelli postulates that siblings act as "models, reinforcers, caretakers,

confidants, pace-setters as well as teachers" (1976, p. 100) and suggests that peers can

also perform these roles effectively. H e considers that important and enduring

commonalties exist between siblings; they experience similar parenting practices,

socialisation patterns are likely to be the same, methods of communicating will be

known to one another, and finally methods employed for the transmission of learning,

culture and knowledge will be shared (Cicirelli, 1976). Conversely though, more recent

research (Dunn & Plomin, 1990) claim that siblings' experiences are very different from

each other, since they in effect operate and interact in quite different environments from

one another.

Nevertheless sibling behaviour is of interest in this investigation for two reasons.

First, it m a y be that toddler participants w h o have older siblings will, because they are

likely to be more practised at symbolic play, display symbolic-play activity more

frequently and at a higher level than toddlers without older siblings. Results of studies

of sibling play in the home (Dunn & Dale, 1984; Dale, 1989; Youngblade & Dunn,

1995) support the expectation for similar findings to occur in this current exploration (to

be undertaken in L D C C C s ) .

Secondly, the third condition (dyad play) in which children will be observed in

this project requires the pairing of the toddler participant with an older familiar

preschool-aged peer. It is planned that an older, familiar and if possible 'self-chosen'

by the toddler participant will be the more-experienced play partner in the dyadic-play

situation. The expectation is that such a partner can be identified in the group mixed-

age free play condition (Condition T w o ) since toddlers often have specific play partners

among the older children. It is surmised that if the older, preschool-aged peers can be

identified in this way (as familiar companions to the toddler) they may, when engaged

54


in play or other interactions with the toddler, display similar behaviours to those of

older siblings.

Support for this notion is gained from the work of Ellis, Rogoff, and Cromer

(1981) w h o reported that mixed-age play was a dominant feature in their naturalistic

study assessing neighbourhood play. (Same-age play was found to be more prominent

when children were seven or eight years old - the later years of early childhood.)

However, the question remains whether or not long-term affiliations between younger

children - infants and toddlers - and older peers in L D C C C s manifest similar reciprocal

interactions as are seen between siblings of a similar age.

Bronfenbrenner (1979), from an ecological perspective, offers some support for

this notion. Siblings and peers are both depicted as belonging in the microsystem level

in Bronfenbrenner's ecological model and so, as components of the child's immediate

surroundings, each are available for engagement by the child in interactions and

activities.

Children in mixed-age grouping in L D C C C s m a y experience a similar

relationship to that of siblings, in that many c o m m o n experiences are shared. Such

shared experience m a y result in what Dunn and Dale (1984, p. 139), with reference to

siblings, term as the "close meshing of each partner's verbal and non-verbal actions."

Thus, when considering the relationships of children in L D C C C s , it is possible that

these day-care 'partners' would share considerable mutual social and emotional

experiences. This expectation though, is tempered by Dunn and Plomin's (1990)

considerations of siblings' differential experiences.

The research investigating the influence of older siblings on the development of

their younger siblings' play behaviours is presented below. As before, particular note is

made of the younger siblings' imitative behaviour and the older siblings' supportive

behaviour, since each is thought to be an influential component in the present research

project. The studies thought to be most relevant to the current project are listed in Table

2.5. A discussion of each follows the Table.

55

•

3 ca

H

OJO

a a -a w Ml S 9 O >H

u

•ss ca o CO

W)

ca a 55 •s u fl

yy

A HH

5 OJD

fl

•a -mm.

a

a w 3 2 ey

CO

60 8

1

I CO

ea 4)

F-H

X! ea

'5 > 4-P

§ X 3 4) CL. 4) X 3

i — <

CO

O yy

4) to

60S

z <1

60

a *4-4

4> CO

Author(s)

Date

Older sibling offered younger

child support and

assistance in dealing with environment

and play

materials.

The 'monitoring' of older siblings behaviour by

younger children.

Interactions (Eckerman

et al. 1975; Lamb, 1976b,

1977a)

Mother -toddler,

father-toddler and

sibling - toddler

dyads

24 min.

(8 with mother,

8 with father,

8 with sibling)

-et CXI CN y

Laboratory

playroom

Lamb

(1978)

Younger siblings increased activity and

exploration in part

Due to the opportunity to observe and then imitate

their older siblings behaviour.

Infant locomotion

exploration (Author's

scale)

CO

X ea >-, X 60 .3

3

3

•a o X CN

-* r^ oo — i — i < N

Unfamiliar

garden

Samuels

(1980)

When younger

siblings were engaged

in symbolic

play with their preschool-aged siblings, everyday

routines and household

activities were

the most

common themes.

Younger children imitated the older siblings'

pretend actions or comments.

Maternal

interview re

sibling-child relationships

(Authors' scheme)

Joint pretend play

(Dale, 1983)

Classification of play

themes (Dunn & Dale,

1984)

Older preschool-aged

sibling

3 o .3 yy

X

O vo CN CN

s O

X

Dunn &

Dale

(1984)

CO 60 3 • I — I

X 3 yU

CL 3 yy Ul

B £ ea 4_p u CO ea u

ftia 42 "g fi § to B

X ,ea co .S

co ea O 3

-i

u ea u "3 O CL

S£ 2 ,=a

ea 4-p , p 3

l-p .y O 4) 4-P (J

^ s 2 4-1 .-H fl Eco o

4-4 00

o X -1

•? §"8

s a§ ;H ft(S tu o y

3 o "^ 4)

4) Up O

>• 4) ca —^

3 x * 4)

X E 4) 4-P

D.

3

CL

E o u

4)

o eo «a ea tj 60 =a 3 -a 4) g co ca ca x P, 3

o 2, tj cr e3 P PL, cfc

60 3

X 'to

*3 % X^

•g ca

B a XI

rS

a X

3 .fe 4)

5

O X

4)

'>

4)

3 4-1 ca Up

4)

O

3 4> 3 O

I O

o •s o x Up

c2

I CO ea 4)

frl^e B r-

x 4) X

o U ea

co j) co - 4>

x 3 x % ea CL 3 g CO § 8-lffi

CO, 4)

x E 6 2 2 -5 •" u yy-y

3 c

3 ^ ^ 3

£* 3

.8 .2- g

Tl 9.6

3 4)

P

42 3. o 3 x E CO 3

X u E O

3 4) X B _4) 4) "3

&| "s ^ y >

Xi

o

CO X

4) ea bO >p ea X

o 4) ea 60 ca

s > CO •s ^

p£vs:5 4) X

PL, X .25 to >

B G

4) 60 ea

X

ea

X r-1 co 4-1 . •S 60 •g 60

'fe "S -S ? .S ? 3 X 4) X

„ u ea X E X

fe 2 irs 'ffi ° '^ 4) X 4)

W yy ^

60 o x 12 x a -s a

g 2 l o §0

X

o

Q O

X

<N

p-

x CN

8 4)

to 4) CO CO ca CO 3

3 4)

4) O B 3 yy

G

4)

vo" CN

e3 CL, CO 3

g '> CL 3

X 4)

1 U 3

4) JS

box < fi

Ov VO

>n v, CN r-r—( .—<

X 4) to CO 4) to

X^ ea

3 3 •xi 2

:s 1 o m o

60 3

•a 4-1 4) CO

I o X

1 IS

o X <U

3 ta

< O

x 8 O « joSS-•a "n °°

a £ o\ H O C *

-2 li

l> -8 3 )Q 3 3 Ov O 3 Ov

>- Q C

>1 X 3

to

2 o E 4)

Q *


Lamb's (1978) work with 18-month-olds and their preschool-aged siblings

illustrates the input and choice exercised by a younger child when interacting with an

older sibling. L a m b observed the sibling dyads operating in a laboratory playroom

under two conditions. In the first condition the siblings were alone, while the second

condition required either one or both parents to be present. The parents had been

requested to be responsive to initiations from their offspring but not to initiate

interactions with or between the children. W h e n the infant and sibling were alone, L a m b

(1978) observed that the younger children monitored their sibling's presence, actions or

behaviours, but if either or both parents were present the younger child showed a

preference for interacting with them to the detriment of play with the sibling. This

occurred despite the latter's demonstration of affiliative behaviour toward the toddler.

L a m b (1978) noted that, in terms of broadening the toddler's experiences, the

sibling offered support and assistance in dealing with play materials and the

environment. H e suggested this was in contrast to parent-child interactions observed by

him in earlier research (1977b). These interactions L a m b categorised as mostly directed

and initiated by parents.

In relation to the current research, what is most interesting about Lamb's (1978)

study is the amount of 'looking' behaviour, both at the sibling and his or her activity, and

the imitative behaviour engaged in by the infant in respect of the older sibling's actions.

Infants and siblings were observed alone together for periods of eight minutes. In that

period the infant looked at the sibling twice as many times as the sibling looked at the

infant - accounting for almost 25 percent of the infant's interactive behaviours. The

infant also looked at the sibling's behaviour or activity and this was scored separately

from the 'mere' looking-at-sibling behaviour. The two looking behaviours combined

represented more than half the infant's total interactive behaviours (Lamb, 1978).

These results appear to support the idea that the younger child is instrumental

not only in ensuring proximity to the more-skilled other through constant 'looking'

behaviours, but also in operating in some type of affiliative, reciprocal capacity through

engaging in imitative behaviours. Such observations give credence to the notions of

'tutor-proneness' proposed by Bruner (1974, p. 697) and lend support to the Vygotskian

(1978) notion of the importance of the 'more-skilled other' in the acquisition of learning

or new skills. Had the older sibling not been present and the infant been either alone or

with a same-age peer, it is likely that the researchers would have recorded very different

behaviours in terms of social content. However, this conjecturing does not explain why

58


the younger children preferred to interact with their parent or parents when a choice

existed.

In 1980, Samuels published the findings of a study investigating the influence of

older siblings on their younger siblings' exploration of an outdoor environment. The

younger children (toddlers) ranged in age from 17 to 28 months, the older siblings'

average age being four years and two months. T w o twenty-minute observations of the

younger children were made both in the same unfamiliar garden setting with the

mothers present on each occasion. In one observation the toddlers' older siblings were

present; in the second, the older children were absent (Samuels, 1980).

If the toddlers' older brothers and sisters were in the environment, the younger

children increased their motor exploration together with showing more interest and

exploratory behaviour in other aspects and objects present in the garden. The toddlers

also went further away from their mothers when their siblings were in the garden.

Samuels (1980) surmises that, in keeping with the findings of other researchers,

including the work of L a m b (1978) cited above, the toddlers' increased activity and

exploration is in part due to the opportunity to observe and then imitate their older

siblings' behaviour.

The work of Samuels (1980) supports the notion of younger siblings' benefiting,

in terms of their learning and development, from time spent with their older siblings.

Children in L D C C C s have daily exposure to each other in a shared and constant social

environment. The notion, that in L D C C C s relationships between differently aged

children m a y come to resemble those of siblings, is persuasive.

Symbolic play was just one area of interest for Dunn and Dale (1984) when they

observed 26-month-old children in their own homes, under the following three

conditions: playing alone, playing with their mother and playing with their older sibling.

Participants were observed for an hour in each of the three conditions. Several findings

resulting from this study are of interest to the current investigation.

First, the older sibling established symbolic play and then invited the younger

child to join in. The resultant symbolic play required the players to engage in

"transformations of role, identity, location and psychological state" (Dunn & Dale,

1984, p. 136). Of interest to the current researcher were the younger children's

initiations to their siblings - more than one third of this involved imitation of the older

child's actions or comments. Finally, close to one third of the play with siblings relied,

for continuity, not upon objects or props, but on language and non-verbal actions. Each

59


of these findings lends support to the current research. As previously mentioned it is

expected that the younger peers' will engage in imitative behaviour in Conditions T w o

and Three (the conditions in which play with older peers occurs). Dunn and Dale

(1984) recognised that siblings enjoy a shared context containing c o m m o n experiences

and activities, which would support their engagement in symbolic play. It is postulated

that children attending L D C C C s also experience a c o m m o n context of experiences and

activities - commonalities which are likely to support their shared symbolic play.

However, in accordance with Howes and Farver (1987) and Rothstein-Fisch and Howes

(1988), w h o also pondered the notion of peer familiarity in L D C C C s being akin to that

of a sibling relationship, it is recognised that a difference exists between the emotional

intensity of a peer relationship and that of a sibling relationship.

Dale's (1989) interest in the facilitation of a younger sibling's symbolic play by

an older sibling continued. In a later study she observed toddlers aged 2 years old with

their 4- or 5- year old partners engaged in play on two occasions, each of one hour's

duration, in their own home. Dale (1989) noted the following: more than half (59

percent) of the younger child's pretend play was conducted alone; a quarter of the

pretend play was in conjunction with the older sibling; 16 percent was shared only with

the mother; and the remaining play was with both the mother and the sibling. As in the

earlier work by Dunn and Dale the symbolic play between siblings was more complex

involving "role, location and psychological state transformation" (1984, p. 140).

Comparing the play-styles of mothers, fathers and first-born siblings with infants

and infants playing alone on one hand, and the influence of these styles upon the

infant's cognition on the other, was the objective of research reported on in 1988 by

Teti, Bond, and Gibbs. The infant participants were filmed for 10 to 15 minutes with

their first-born siblings as play partners when the former were aged 12.5 and 17.5

months. The older siblings ranged in age from 28 to 93 months. The observed play

behaviour was coded across seven categories of experiences including spatial,

perceptual, fine-motor mastery, language mastery, object play and gross motor activity.

Differences in play styles were found to exist - for example substantial parallel play

was a feature of the play between the younger infant participants and their first-born

siblings.

Of interest to the present researcher was the finding that first-borns were noted

as responding to, as opposed to influencing, the infant participant's play. Of further

interest was the suggestion that, in response to the infants' 'object skill' at 12 months, a

60


tendency for the older siblings, regardless of their age, to verbally stimulate the infants

was identified.

Youngblade and Dunn (1995) observed young children aged 33 months as they

played and interacted with their mothers and older, preschool-aged siblings. Play under

three conditions was observed - the participant's play with mother, sibling and with

both mother and sibling. T w o 75-minute observations, a week apart, were completed

and the collected data coded for social-symbolic play. Symbolic play was identified

when "actions, persons, objects or places were transformed or treated non-literally"

(Youngblade & Dunn, 1995, p. 1476.) The researchers also noted whether the symbolic

play exhibited by the child featured 'role enactment' (entailing mental representation

only) or 'role play' (requiring true symbolic activity). (While the primary objective of

this research was to identify young children's expertise in understanding other's beliefs

and perspectives, symbolic play was the medium through which these skills were

assessed.) Youngblade and Dunn (1995) reported that the younger children engaged

more frequently in symbolic play with their siblings than with either their mothers or

mothers and siblings together. In addition the younger participants exhibited more

complex play with their siblings than in either of the other two play conditions.

In the present study the symbolic play behaviours of 'role enactment' and 'role

playing' - as noted by Youngblade and Dunn (1995) - parallel components of

Dimension Four symbolic-play behaviour ('self-other relationships') according to

Westby (1991). Such behaviour is aligned with 'decentration' - an increase in which

was found by Lyytinen (1989) for all participants (aged between 2 and 6 years old) in

her research involving dyadic play. (Lyytinen's study was reviewed earlier in this

chapter in part 2.2.)

The findings of these investigations by Dunn and Dale (1984), Dunn (1989) and

Youngblade and Dunn (1995) lend weight to the expectation that toddlers with older

siblings are likely to exhibit symbolic play more frequently and at higher levels than

toddlers without older siblings. It is surmised that toddler participants with older

siblings are likely to be more skilled in symbolic-play behaviour as a result of their

interactions with siblings. It seems reasonable to speculate that such skills will be

transferred to their symbolic-play activities in the L D C C C environment.

61


2.8 LITERATURE EXAMINING SOCIAL INTERACTIONS B E T W E E N

SIBLINGS

The research dealing with sibling interactions - more specifically those between

younger toddler-aged children and their preschool-aged siblings - is also pertinent to the

current study. Once again imitation of the older sibling's behaviour by the younger

sibling is evident - this behaviour is, as mentioned earlier, of particular interest to the

present investigation. Table 2.6 indicates the studies referred to in this section. A

discussion expanding upon the content of each of the studies cited follows the Table.

62


In 1979 Abramovitch, Corter, and Lando reported on the first of two studies to

be reviewed here focusing on the prosocial (positive), agonistic (negative) and imitative

behaviours of siblings with a range of age intervals. In the first project, 34 same-sex

sibling dyads were observed for an hour, on two occasions a month apart, in their own

homes. The difference in the ages between the siblings ranged from 20 to 38 months

and the age of the older sibling ranged from 34 to 64 months. Results indicated that the

younger child in the dyad - aged approximately 20 months - engaged in a large amount

of imitative behaviour some of which had implications for learning skills (Abramovitch

et al., 1979). The researchers determined that, in this first study, there was no effect for

either age interval or sex.

Although this result (incidence of imitation) was not statistically significant, in

terms of the imitation process and its relationship to child development it remains

important. At approximately 17 months toddlers are likely (as a result of their ability to

form mental representations) to experience a surge in deferred imitative behaviour,

which increases in complexity as their symbolic activity increases (Flavell, Miller, &

Miller, 1993).

Abramovitch, Corter, and Pepler continued investigations into sibling dyads and

in 1980 published a study involving mixed-sex sibling dyads. Once more the objective

of the research was to record the positive and negative social behaviours and imitation

observed between the siblings. Again the participant pairs were observed in their own

homes on two occasions and, as before, the length of time between the two observations

was a month. Results of the 1979 study were replicated in that the younger sibling

engaged in substantial imitative behaviour (76 percent of all imitative behaviour

recorded) and as before no effect was found for sex or age of older sibling.

Imitation of older siblings' behaviour, especially that of symbolic play, is

thought likely to be reflected in the play engaged in by toddler participants in L D C C C s .

The findings of each of these studies - Abramovitch et al. (1979, 1980) endorse the

prediction that the skills of toddlers will be enhanced through their imitation of their

older, more-skilled preschool-aged siblings' behaviours.

64


2.9 LITERATURE EXAMINING SOME INFLUENCES OF ADULT

INTERACTION AND TEACHING STYLE

As foreshadowed earlier in the previous section, specialist Early-Childhood

Education training is thought likely to aid staff in their interactions with and planning

for the young children in their care. Professional early-childhood teacher training and

pedagogy involves studies in child development theory and research; consequently

graduate teachers would be aware of the importance of supporting the young child's

developing symbolic play. In an attempt to assess the knowledge and understanding of

this area, a component of the current project is a staff questionnaire the content of which

elicits caregiver attitudes towards play and a number of other features of the early-

childhood day. This instrument also details the level of early-childhood training

achieved by each respondent.

In early-childhood settings - for example, L D C C C s - positive, nurturing and

supporting teacher-child or caregiver-child relationships are viewed as essential to the

child's well being. The degree to which these aspects feature in the adult-child dyad

depends very much upon the teaching style practised by the adult. This section of the

chapter will review the more informative of the research available inquiring into the

nature of teaching styles and teacher interactions. First though, the findings of three

studies reporting on investigations into teacher training are considered. These studies

and others are listed in Table 2.7. A discussion of pertinent issues follows the Table.

65

<u

s es

H

a*

CM

•JB

•a ea

g>

H

1

o

1 3

i O CO

8f •a

9

2

CO

oo 8

1

CO

"a" i ca

4)

•s >

1 1 CL

.2 e1 z-8

„-. eo

z < e

(50

3

00

Author(s)

Date

Training (in this case

in communications

And child development) does effect the

attitudes and behaviour of caregivers

toward young children.

Differences in child-rearing

attitudes

(Parental Modernity Scale,

Schaefer & Edgerton, 1981)

Caregiver interaction

(Author's scale)

Interactions between

caregivers and

children in

LDCC

2x45

min.

59

Adults

U U

u

s

Arnett

(1989)

*A

The two most experienced

adults

engaged in the three most

effective styles

of play interaction.

The two

adults with

less experience used

less effective styles of play interaction.

Teacher play styles

- 6 categories

(Authors' scale)

Safety monitoring

(Authors' scale)

4 types of children's play

behaviour

(Authors' scale)

Teacher-child interaction

play

style

13 hours

(with 3

different

groups

of

children)

4

adults

(classed as

2 'more

experienced'

and 2 'less

experienced')

University

laboratory

preschool

Enz &

Christie

(1997)

*A

"8

3 ea

o, 4-P

a 4)

-S 4)

3 X ca

o 3 4)

X *

60 3 X 8 ••*-<

yy

4)

* CO

ea 4)

yt

60 3

ea h 4>

60

'e y^

O <N

5 a X >

a o X c+2

x -"g

P

3 CL 8 ,-k

• - < . — i

a^ •g *•

S oo' a o\

.S 2 <u -

8=1 .2 « C fa

X 4) X 3

1 'T's 4) — t« 4)

"3 £ 3 «J 3

X

c ea

u I

yy

ea 3 3 O o

a CO 11) 3

a X O ea 41

H

(i)

o > a a C/J

ii

>,

60 « 3 8 •3 4)

,E E « b £ cr s" 2 -

•2 S b 3 2 ^ 3 X 3 •5 x <, 4) O w

— 3 «3 "" 3 ,M

.2 3

es U X X O

a -̂ a x b 3 ° o 2

1s.| 2 x c 2 8 °

.Sou a a u co ea X

o a a 3 co 3 O

4)

B 2 4)

I o

£"23

x a i a 2 «« X -y X ea X o £ H o 3

& S x 2

IS '5 a a"? xO'^ x a "i x fi oo o 2 os ii. s y

w a „r £ « S y ea 3

c3 a s

ca

y ga •2 js « o 2 > o a -̂ eo x ca

2 ° 8 -> X r j X

a 2u-g 2 3 X a w o

"S § a x 2 <u

3

8 60

o u

X o

t; o CL

§• co co 4>

Tt ea y UP y >-> a ca 4) S

a & > >

ca ea

I 60 3

'fi

a

a X 4)

H

x CO 3 O

oo Os

4) C-

X °° « OS

•4—1

4) 00

U

r ea

"y g*J

6 J B

i tw *S-5

o

•8 § X X

CO

' ^ X

e9 <fi

X ea

"S >

* - p

8 ii X 3 4)

a 4>

1 §

u a *? H -fi M

U

u u

3 x .S xi .& V a 5 3 a .2 o a ca ca

H 8

a XI to a-fi

>^x a o "3 3 CT 4)

a 3 4) CJ

a x

B 00

3 °

5 I *$ - 8 — >> 60

8 O to o

I ea

oo

OS Os

CO O x o

TT X

3 o X cs X

2

g fi X a ca

3

9| §x cs | g * X © B «? y&.B?y

4) p>

*-* G 60 O cj

CO CU

a > CJ ca

>1

•s

3 ea

ex 'o •a CL 4) X

4> rj 60 *3

< fi 00

a ? — p

X CO eo X y-l g. Q

J. a "a —' w ea

CO

O 0 X

CJ , co

+ 4) SO &

60 8

a 4> •V3

U

u u

a

u u u Q

u u u Q

3 X ca CO

2 o 3 4) X

I < *

o •S 2 3 ea

< Q

nig&

allal

98)

O .3 OS

X X z. < *

OS OS

=3 CO y. ^y.

P 3 u~> O 8 OS

B w C

CO

60 3 8 y-,

CO v '

4)

% CO

ca

a

B X ea

> G 4) X

g &

ss-8

8>£ a < e

60 3

"3 4) 00

Author(s)

Date

Children engaged

in more meaningful

and developmentally

critical

behaviours when interacting with

adults who are "responsive,

supportive and

relatively

nondirective" (Mahoney & Wheeden,

1999, p. 67).

Teacher behaviour scale

(Mahoney, 1992 adaptation of

Maternal Behaviour Rating

Scale, Mahoney et al., 1986)

Influence of teacher's

interactive style on

child's persistence and

involvement in

activities

3x7 min.

1x7 min.

alone

2x7 min.

with teacher

49

actual ages

17-71

(development-

tal levels 8-39)

Preschool

s Mahoney &

Wheeden

(1999)

A positive association between

teacher-

child interaction a and

toddlers' social bids.

ECERS

(Harms & Clifford, 1980)

ITERS

(Harms, Cryer, & Clifford,

(1990)

Predictors of social

outcomes

for children

1x8 hours

(one day)

718

11-18 (infants)

18-36

(toddlers)

36-62

(preschoolers)

U

u u Q yy

McCartney

etal.

(1997)


A n important consideration for research involving L D C C C staff is the level of

training achieved in Early-Childhood Education. This m a y range from a complete

absence of formal training to a degree or higher degree in Early-Childhood Education.

Arnett in 1989 published the results of an investigation into whether training

makes a difference to care-giver attitudes towards and interactions with children. T w o

different observers rated 59 staff for 45 minutes on two separate occasions. Staff also

completed a questionnaire designed to assess authoritarianism. Arnett's findings clearly

indicated that training made a difference to both attitudes and behaviour. Trained staff

were less authoritarian and displayed more positive interaction behaviours (1989, p.

549). As his investigation was linked to a specific training program, he was able to

determine that completion of the first year of the training program changed participants'

attitudes and behaviours. Of interest to the current researcher is that the units studied in

this initial year of training were communications and child development theory and

research. Each of these would be likely to affect participants' view, first of the nature of

children and childhood, and secondly of appropriate methods of interacting with young

children.

More recent research supports Arnett's findings. Enz and Christie (1997)

examined teacher play interactions styles and the effect of these on children's play

behaviours. Three play styles from the six identified by the researchers were most

suited to supporting and guiding children's play - including that of symbolic play.

Importantly these three styles, engaged in by the two highly trained and experienced

teachers, ensured the control of the play remained with the players. Of concern was the

finding that the two less-experienced staff used play interaction styles that could impede

rather than effectively support young children's play. Enz and Christie (1997)

concluded that both training and experience affected teachers' choice of play interaction

strategies.

Honig and Hirallal (1998) first observed 81 teachers in the preschool classrooms

of L D C C C s and then secondly, collated their responses to a questionnaire describing the

years of service, level of education and early-childhood study achieved by each.

Analysis of the data showed that training in Early-Childhood Education and child

development accounted for 62 percent of the variance in teacher inputs reflecting

positive teacher interactions.

Teacher education was also a component of a study published by McCartney et

al. in 1997. Their research focused on investigating a number of factors, including

69


teacher-child interaction, thought to influence social outcomes for children in L D C C .

Findings indicated that the number of social bids made by toddlers, aged from 11 to 18

months, was positively associated with teacher-child interaction. In addition, both

teacher income and teacher education were significantly and positively associated with

teacher-child interaction, lending support to the importance of training in Early-

Childhood Education for L D C C staff.

Studies indicating the importance of staff training are relevant to the current

project it is expected that staff trained in Early-Childhood Education have a detailed

knowledge of child-development theory and research. Such knowledge and

understanding enables staff to offer developmentally appropriate provisions and support

for the children in their care in order for them to practise and enhance their symbolic-

play skills. Staff w h o participating in the current study were asked to complete a

questionnaire identifying the level of training in Early-Childhood Education they have

achieved. It is predicted that the toddler participants in L D C C C s where the teachers are

trained in Early-Childhood Education will exhibit symbolic play more frequently and at

a more sophisticated level.

Teaching style and mode of interaction are behaviours that impact on the

teacher-child or caregiver-child relationship and are thought to influence the child's

experience in care. Each has warranted investigation. Howes, Hamilton, and Matheson

concluded, after a three-year study, that a positive 'first' teacher-child emotional

relationship namely, emotional security - the authors suggest this is "close to the

attachment construct of security" (1994, p. 261) - promotes social competence with

peers. These findings were the result of a project, conducted across six centres,

involving 72 children aged, at the start of the study, from 13 to 24 months.

Maintaining this focus Howes and Smith (1995) explored children's playful and

cognitive activities in terms of their relationships with their teachers and other

dimensions of the care-giving environment were also taken into account. Eight hundred

and forty-two children, aged either between 10 and 13 months or 36 months or over,

were drawn from 150 L D C C C s . Findings confirmed that those participants w h o were

emotionally secure as a result of the teacher's positive social interactions were also

engaged in more exploratory, creative and cognitive activity.

Current findings by Mahoney and Wheeden (1999) support those of Howes et al.

(1994) and Howes and Smith (1995). Mahoney and Wheeden investigated the teaching

style of staff working with preschoolers with special needs. While the 49 participants'

70


chronological ages ranged from 17 to 71 months, their developmental range was from 8

to 39 months. Observations were made of children engaged in solitary play, teacher-

child dyads in free play and finally a condition whereby children were instructed by the

teacher on how to perform a play activity. The objectives included assessing the

children's attention, persistence and involvement in activities. Mahoney and

Wheeden's findings indicated that while, on their own, children engaged in activity

likely to support their development, with adults w h o were "responsive, supportive and

relatively nondirective" (1999, p. 67) children were likely to "initiate meaningful and

perhaps more developmentally critical behaviours" (1999, p. 67). It seems adults'

affective supportive responses encourage children's initiation of behaviours likely to be

significant in their development.

It is expected that staff trained in Early-Childhood Education are likely to be

more supportive of toddlers' emerging symbolic play skills than those who are

untrained in this profession. The studies undertaken by Howes et al. (1994), Howes and

Smith (1995) and Mahoney and Wheeden, (1999) are important in terms of this

expectation and the current project from two perspectives. First the findings reveal that

teachers' interactions or teaching style m a y be 'responsive', 'socially positive',

'supportive', 'non-directive' and 'emotionally positive'. These descriptors are

synonymous with those behaviours supporting young children's secure attachment. In

addition, positive teacher-child interactions were found to support children's

competence with peers.

A central component of the parent-child relationship is that of attachment and an

abundance of evidence illustrates the association between those children identified as

'securely-attached' and symbolic play ability (Slade, 1987a, 1987b; Meins, Fernyhough,

Russell, & Clark-Carter, 1998) and positive peer relationships (Kerns & Barth, 1995;

Pastor, 1981). Each of the studies cited in this section suggests that components of the

relationships between teachers and young children attending L D C C C s reflect, in part,

those of the mother-child relationship. Regularly attending children m a y be expected to

establish secondary secure attachments to their caregivers, and these subsequent close

and intimate relationships would be likely to support both their exploration and learning

and the development of peer relationships.

It is argued that staff, trained in Early-Childhood Education would, as mentioned

earlier in this section, first be more cognisant of the importance for young children to

71


develop symbolic activity and secondly be aware of the need for young children to

establish secure trusting relationships with their caregivers in L D C C .

2.10 LITERATURE EXAMINING SOME INFLUENCES OF THE DAY-CARE

ENVIRONMENT ON TODDLERS' PLAY

Of relevance also, to this study, is the literature reporting on the situational

variables present in the L D C C environment and the influence of these on toddlers

involved in symbolic play. Johnson and Ershler (1982) recognised the importance of

identifying and understanding both the curricula and contextual effects on play and

proposed that it was useful to identify the "mediating variables and mechanisms

responsible for them" (1982, p. 137). It would seem likely that contextual variations,

reflecting the quality of practice or provision of care in L D C C C s , might enhance or

depress both the level and quality of play outcomes.

In Australia, quality in L D C C C s refers to a range of factors in the physical and

social environment, aspects of which must comply with a nationally specified standard

- modelled on that implemented by the National Association for the Education of

Young Children in America ( N A E Y C ) . The notion of a high-quality L D C C C suggests

a well-resourced physical and social environment offering professional,

developmentally appropriate provision of care for young children.

In accordance with other parts of this literature review the impact of the quality

of the early-childhood environment is explored with reference first to the development

of young children's play and secondly to young children's social interactions. Social

competency, as previously mentioned, is a component or precursor to successful social

play activity.

The first range of literature reviewed in this part of the chapter explores the

influence of a quality day-care environment on a number of aspects of young children's

play development. A brief discussion of each research project follows Table 2.8.

72

CO

60 8 X B yy

CO

•.—'

1 CO

ca 4) ty 4)

'S > fi 4) X

g ex 4>

"S p — p

CO

X O

« X 60 X!

fc < fi

60 B

•a 4) GO

Author(s)

Date

Quality of care and family

characteristics positively influence

children's play with adults, toys and

peers.

Quality FDCRS

(Harms &

Cifford,

1984)

Play with adults scale

(Rubenstein &

Howes, 1979)

Play with peers - Peer

Play Scale (Howes,

1980)

Play with objects

scale (Rubenstein &

Howes, 1979)

Family

and child-care

influences

2x15

min,

55

11-30

S U >> ea

Q

1 ea

yy

Howes & Stewart

(1987)

Children were more social and had more

knowledge of problem-solving

in high

quality settings

ECOI (Bredekamp,

1985)

Free play

Observations

Problem-solving

procedure (Spivack &

Shure, 1974)

Quality of LDCCCs

and preschools

e

o CM X CN

55

48-59

LDCCC

Preschools

Holloway &

Reichhart-

Erickson

(1988)

CO

60 8 X 8

E

CO

"of 5 CO

ea 4)

yy

Ji IB ca

> -p-J

3 4)

•g 4) Q. 4) X 8

1—<

C O

X

o fc

p . - M

fc < fi

60 8

a 4)

c/3

Author(s)

Date

Children attending long-day child-care

engaged in more cooperative play.

Child's behaviour

(Italian adaptation of

Oxford Preschool

Research Program,

Bruner, 1980; Sylva,

Roy &

Painter, 1982;

by Amaioni, Bascetta

& Aureli, 1988)

Early child-rearing (home

or LDCCC)

E*

X

o CN

40

38-42

Preschools

(Italy)

Aureli &

Procacci

(1992)

Children experiencing good quality

care engaged

in more complex play

more frequently.

Quality ECERS

(Harms &

Clifford,

1980)

Play

a) Peer Play Scale

(Howes, 1980)

b) Social-Pretend Play

Scale (Howes et al.

1989)

E

§ a, 2 13 > 4)

a, y.

ii

PU

6x4x2

hour

259

10-15

CO

u u u

3

Howes &

Matheson

(1992)

Study 2.

FDC children showed more pro-social

behaviours and higher quality play

behaviours.

Influence (effect) of

home, or FDC, or

LDCCC setting on

interactions and play

behaviour

VO UP

o

18

24-36

Home

FDC

LDCCC

Lamb,

Sternberg, Knot,

Hwang, &

Broberg

(1994)

CO

60 8 X 3

E

CA

I CO

ea 4)

JJ X ea

'S > 4-1

B 4) X 8 4)

ex 4) X 3 yy

CO

X o fc

4) .£ 60 Xi

fc < B

60 ,3

a CO

Author(s)

Date

Children with LDCC experience

engaged in more complex symbolic

play and more peer interaction than

those who had not attended LDCC.

Cognitive play

(Smilansky, 1968)

Cognitive Level and

Thematic Unit

(Aureli & Colecchia,

1996)

Social play (Smith's,

1978 reduced version

of Parten's, 1932

Scale)

Early child-rearing

(home or LDCCC)

20x1

min.

O vo rj- en

CO

"o O X CJ CO

PL,

Aureli &

Colecchia (1996)


A number of studies offer insight into the relationship between high-quality care

and children's play and have proved useful in informing the current project. The first,

reported on by Howes and Stewart in 1987 involved 55 participants aged from 11 to 30

months attending Family Day Care settings. Each child was observed on two separate

occasions for 15 minutes. The dependent variable was children's play with adults, toys

and peers as influenced by the family and child-care influences. Findings revealed that

each of these (quality of care and the family's characteristics) influenced the

participants' play.

The researchers found a difference was discernible between the choice of

L D C C C by families who were nurturing and supported and those who were stressed -

the latter often choosing unstable, low-quality child-care arrangements, whilst the

former accessed and retained higher-quality arrangements for their children (Howes &

Stewart, 1987). Relationships were found between nurturing, supportive family

conditions, stable, high-quality child-care and positive outcomes for the participants'

play development.

Holloway and Reichhart-Erickson in 1988 published the findings of an

investigation into free-play behaviour and social problem-solving skills of 55, four-

year-old children. (The following year they published the results of their exploration of

links between peer relationship skills and maternal expectations - this is reviewed in the

next part of this chapter.) Both studies were undertaken in L D C C C s and preschools. In

the first study (1988), teacher training and low adult-to-child ratios were found to be

supportive of young children's pro-social behaviours. (Unfortunately the report does

not clarify what proportion of these pro-social behaviours represents social play

activity.) Dimensions of quality included a 'cluster' reflecting positive teaching styles -

described by the researchers as "respectful, engaging, responsive and democratic"

(Holloway & Reichhart-Erickson, 1988, p. 49). Perhaps not surprisingly these same

teachers were observed to provide ranges of developmentally appropriate play

materials. However, interestingly, in settings reflecting components regarded as

supportive of good quality - adequate space, safe well provisioned areas - many

children engaged in more solitary play and less observing behaviours.

The third study in this section is that in which Aureli and Procacci (1992)

compared the social competency - in terms of play - of two groups of children aged

from 38 to 42 months in their first year of nursery school (the Italian equivalent of the

Australian preschool). Half the participants (20) had attended L D C C C s for at least two

76


years before going to nursery school, while the remaining 20 children had been cared

for by their families until their commencement at nursery school. Those participants

who had experienced L D C C exhibited a range of peer competencies, but most

importantly for this current study, they engaged in more cooperative play than those

children lacking experience in the L D C C C environment. While the centres were not

rated for quality of care, the researchers Aureli and Colecchia (1996) emphasise that

L D C C C s in Italy, whether private or not, conform to parameters - for example, stable

staff and an adult-to-child ratio of one to five - regarded as desirable for quality care.

Italian L D C C C s are regulated in strict accordance with guidelines ensuring the

provision of high-quality care.

The work of Howes and Stewart (1987), Holloway and Reichhart-Erickson

(1988) and Aureli and Procacci (1992) supports the belief that the quality of care

provided by the L D C C C may be an influence on the development of toddlers' symbolic

play. The results of each study show that both children's competencies with peers and

their social-play skills appear to be enhanced by experience in high-quality early-

childhood settings.

While there appears to have been extensive investigation into the issue of quality

in L D C C and a number of aspects of play development, little research has examined the

impact of quality on the development of children's symbolic play. A project undertaken

by Howes and Matheson (1992) has in part addressed this need. Their study was the

second in a series of two projects designed to identify sequences in the development of

what the investigators, Howes and Matheson described as "competent play" (1992, p.

961) in the areas of social behaviours and social-pretend play. One sample of children

was observed in L D C C C s where group sizes and supervision (caregiver-to-child ratios)

were not optimal and above those recognised as appropriate for quality care-giving.

Participants in the second sample attended a 'model' centre where caregiver-to-child

ratios were considered optimal for quality child-care. The observers recorded the

children's highest levels of play development and its frequency. Results, which showed

social-pretend play increasing in both complexity and frequency with age, replicated

those of the first study. The ages at which the various types of play developed for

children in sample two, the centre rated highest in terms of quality, were close to those

in the first study. However, Howes and Matheson (1992) identified a time lag or delay,

in the age that children, in the first sample, (those in low-quality care) engaged in the

various types of social-pretend play. Children, in this sample, engaged in

77


complementary and reciprocal play three months later than similarly aged children in

the first study. The time lag increased for these children's participation in cooperative,

social-pretend play to between 12 to 18 months. Of even more interest and concern is

the researchers' finding that, not only was the development of complex social-pretend

play delayed by at least seven months, but in some cases such play was not exhibited at

all by children attending the low-quality L D C C C s .

A point of specific interest is that Howes and Matheson observed children "only

when free to interact with both adults and peers and not engaged in a structured,

teacher-led activity," (1992, p. 973). This study omitted reference to such variables as

the type of materials and equipment available for the children or the programming

policy of the centre. These components may have had some effect on the outcomes -

for example if appropriate materials supporting the development of symbolic play are

not available for children then other types of play will be engaged in. Details of

programming policy are useful, for example, to identify whether or not spans of

uninterrupted time are available for children to become successfully engaged in play

episodes.

Nevertheless, the results of this study are important for a number of reasons. In

terms of Bronfenbrenner's (1979) ecological theory, L D C C C s feature in the

microsystem of the ecological model because they are a component of the child's

immediate environment. The study conducted by Howes and Matheson (1992)

indicates that low-quality care can affect both the social and play development of young

children. Low-quality indicators include high child-to-caregiver ratios, high staff

turnover and often, partially as a consequence of these factors, a transient child

population. Exponents of ecological theory contend that the environment (both physical

and emotional) is instrumental in providing the stability, predictability and consistency

necessary for young children to flourish. Bronfenbrenner (1979) refers explicitly to the

reciprocity of interaction among the adults and children within an environment resulting

in a closeness or intimacy between those persons necessary to promote the successful

development of the child. Poor or low-quality L D C C C s do not provide this supportive,

nurturing context.

Of additional interest in the Howes and Matheson (1992) study are the ages at

which children engaged in the various types of play. Previous research on play

development had resulted in some consensus regarding the proposed ages at which

different categories of social play emerges (Garvey, 1990; Howes, 1988). However,

78


Howes and Matheson (1992) observed that, for more than half the participants, the

emergence of complementary and reciprocal play occurred at 13 to 15 months, and

more than half the children aged 30 to 35 months had exhibited cooperative social-

pretend play. Complex social-pretend play was engaged in by almost half the children

aged 42 to 47 months. Importantly, children who developed complementary and

reciprocal play earlier than would have been expected also engaged in cooperative

social-pretend play and complex social play earlier than previous research indicated.

Nevertheless, while it is reasonable to expect variation in the development of pretend

play with different samples, it is important to recognise the impact that variations in

quality appears to have on this development.

The children for w h o m complementary and reciprocal play emerged earlier (that

is, between 13 and 23 months of age) were, at 30 to 35 months, rated by observers as

more prosocial and reported by their teachers to be first, less aggressive and secondly to

experience less difficulty with peer interactions (Howes & Matheson, 1992).

The above findings suggest that for some children a good quality L D C C C

experience may well enhance social-play development. If this is apparent in a same-age

peer group perhaps younger children engaging in social-pretend play with older peers,

as hypothesised in the present study, may enhance a greater variety of social and play

skills.

Research by Lamb, Sternberg, Knuth, Hwang, and Broberg (1994) conducted in

Sweden illustrates the notion of children's play behaviours differing according to

environments. The study undertaken by L a m b et al. (1994) required children to be

observed in their own home and also in either Family Day Care or L D C C C s , the focus

being on their peer interactions and play behaviours. It was revealed that, when

children were in their own homes, no differences existed in these behaviours for those

children w h o attended either L D C C C s or Family Day Care. However, in the 'away-

from-home' environments differences were noted in that children in Family Day Care

settings exhibited more positive social behaviours and higher quality play than those

attending L D C C C s . It might be that a degree of familiarity arising from the smallness

of the Family Day Care group combined with the range of differently-aged peers

available for social interaction and play m a y account for this.

In 1996, Aureli and Colecchia published work arising from a different analysis

of the data referred to in an earlier paper (Aureli & Procacci, 1992). In this second

analysis they examined the play (including symbolic play) of children aged three years

79


old as they entered preschool. While children in both groups (20 had attended L D C C C s

and 20 had experienced care only at home) engaged in symbolic play, those participants

who had experienced L D C C exhibited more complex symbolic play and maintained this

type of play for longer periods of time than those children w h o had been in home care.

In addition, twice as many participants in the 'LDCC experience' group exhibited

social-pretend play more than the other children.

In conclusion, the work of Howes and Stewart (1987), Holloway and Reichhart-

Erickson (1988), Aureli and Procacci (1992), and Aureli and Colecchia (1996) lend

support to one of the additional research questions posed in the current study - namely

whether the L D C C C environment (which includes the quality of care provided, the

program of activities implemented, the materials and equipment provided, staff training

and qualifications and other available resources) supports the development of toddlers'

symbolic play. The results of each study reviewed in this section show that both

children's competencies with peers and their social-play skills appear enhanced by

experience in high-quality early-childhood settings.

2.11 LITERATURE EXAMINING SOME OF THE INFLUENCES OF THE

DAY-CARE ENVIRONMENT ON PEER INTERACTION

As mentioned earlier, there exists a paucity of research examining the links

between the quality of the L D C C C and children's symbolic play - a largely cognitive

skill initially manifested through solitary play. The development of more complex

symbolic play involving other players emerges as social competence increases. Studies

examining the factor of L D C C C quality in terms of outcomes for children's social

competencies are therefore relevant to the current study. The most informative are

presented below in Table 2.9. A discussion of each follows the Table.

80

CO

60 8 X 8 yy

W3

5 CO

cd 4)

s

4)

1 'S >

s 4) -3 4)

la's 3>

>— , CO

X O

» X 603

z < a

60 8 •a V)

Author(s)

Date

Children

in low-quality

LDCCCs spent

less time

interacting with peers, those

in

high-quality LDCCCs

experienced more positive

interactions with staff.

Quality questionnaire &

observations (Unspecified)

Range of behaviours towards

adults and peers (Unspecified )

High, moderate and low

quality LDCCCs

a

SO y-i

X y-t

55

36-48

U u u

9

Vandell &

Powers

(1983)

A relationship exists between

quality of

setting, maternal

expectations and children's

social competence.

Social problem

solving

skills

adaptation of scale (Spivack &

Shure, 1974)

Peer

interaction observation

Teacher rating competence

scale

Quality ECCOS (Bredekamp,

1985,1986)

Maternal expectations

adaptation of scale (Hess

et al.,

1980)

Quality of Preschool or

LDCCC, Family structures

and Maternal expectations

for development of social

and cognitive

skills

.8

O CN X CM

55

48-59

Preschool

LDCCC

Holloway &

Reichhart-

Erickson

(1989)

Group size in conjunction with

provision of materials and

classroom organisation supports

peer competencies.

ECERS (Harms &

Clifford,

1980)

ITERS (Harms &

Clifford,

1986,1987)

Q-Set, Waters & Deanne (1985)

Peer

play (Howes, 1980)

Group sizes and adult-to-

child ratios

1 x 2 to 3

hours over 3

days

414

68x 14-

24

175 x

25-36

171 x

37-54

U U U Q yy

Howes,

Phillips, &

Whitebook

(1992)


Competencies investigated include social adjustment, sociability and complex

social play (Vandell & Powers, 1983; Holloway & Reichhart-Erikson, 1989; Howes,

Phillips, & Whitebook, 1992). Researchers' attention has largely been captured by two

quality components of the L D C C C program - ratios (the number of staff to children) -

and group size. Vandell and Powers (1983), investigating young children's free-play

activities, found that in high-quality centres, their participants were more often engaged

with adults in contrast to the wandering and solitary play behaviour exhibited by those

children in centres featuring low-quality care. However, peer interactions were found

not to vary according to the quality of care.

In the later work of Holloway and Reichhart-Erickson (1989) relationships were

found between parental expectations, socially competent children and quality of child-

care provision. The researchers found that socially competent children attended high-

quality L D C C C s and in addition mothers of high socio-economic status and who

expected their children to acquire skills earlier than would be expected used higher

quality L D C C C s than mothers of lower socio-economic status with fewer expectations

for their children's skill acquisition. These findings arise from correlational data so

reflect relationships only. It may be that the high-quality L D C C C s were established in

higher-income areas where children received more assistance at home.

Howes et al, (1992) took as their research objective the exploration of what they

termed as the 'pathway' from group size and adult-to-child ratios to the participants'

social competence with peers. Their study involved observing more than 400 children,

ranging in age from 14 to 54 months, for between two and three hours usually over two

days, in their L D C C C .

Findings showed that favourable group sizes (as opposed to adult-to-child ratios

which m a y not define a favourable group size) in conjunction with the provision of

appropriate toys, materials and equipment, together with classroom organisation,

support children's acquisition of peer competencies. Such competencies include

sociability and social play skills. In the present study, group sizes and subsequent adult-

to-child ratios were recorded at the beginning of each free-play session - as was the

provision of all toys, materials and equipment.

Dunn's (1993) review of these components of L D C C reveals inconclusive

findings relating to the development of social competencies. While at times it seemed

ratios were related to children's development if family factors are controlled for, the

negative impact of low staff-to-children ratios is limited. Perhaps as indicated later in

82


Dunn (1993) what appears more critical, is the total group size. Nine of the eleven

studies Dunn (1993) reviewed found group size influential in predicting the

development of children's social skills. Participation in smaller groups during the child-

care day enabled children to experience positive social outcomes. The reverse was also

evident - large group sizes detracted from children's positive social engagement with

others. Group size is of interest in the current enquiry. The centres in which the present

study will be conducted adhere to the staff-child ratios required by law. A n anomaly

exists though, in that, while ratio regulations determine the overall size of the group,

this size increases and decreases according to the staffs' involvement in the intensive,

'one-to-one' care required by toddlers. Routine care-giving procedures form a large part

of the toddler's child-care day and result in discontinuity in terms of their engagement in

social interactions with their peers and other staff.

2.12 CONCLUSIONS

Areas arising from this section of the literature review identified for specific

consideration in the current project include: choice of early-childhood setting for

research purposes; observed conditions; length and frequency of observations;

programming for mixed-age groups in early childhood settings; age range of children in

mixed-age groups; selection of older-peer partner; and caregiver or teacher influence.

Older-peer communicative behaviour and the toddler participants' imitative behaviour

were two other areas arising from the literature review noted as likely to warrant close

observation. Each of these will now be addressed.

2.12.1 Choice of Early-Childhood Setting for Research Purposes

While access to L D C C C s attached to a university was available for the current

project, it was evident that the demographic features of a centre aligned with a

university might not be so consistently present in centres in the wider community. It

seems useful therefore that this project be conducted in L D C C C s in the wider public

environment. Stratified random sampling of L D C C C s within nine kilometres (or

approximately five miles) of the Sydney Central Business District was used to ensure a

broader demographic representation.

83


2.12.2 Observed Conditions

As mentioned earlier, in the present project toddlers are to be observed on three

different occasions, in the following three conditions: same-age play, mixed-age play

and dyad play - a total of nine observations each of 10 minutes duration.

Guidance for observing the participants under these conditions is taken from the

work of Lougee et al. (1977) in a number of ways. The first was that the younger

child's attention to the play activity increased in the mixed-age dyad. While this finding

may certainly be related to the child's stage of development, it nevertheless accords with

the expectations in the current study. Secondly, the participants' (both younger and

older) ability to adjust their social behaviours to accommodate the others' demands

informs both the current investigation and the research design.

The work of Howes and Farver (1987) lent support and guidance to the

construction of the study reported on in this thesis in two main ways. First, in response

to the findings by Howes and Farver (1987), one could argue that it is likely that more

than one period of mixed-age free play (either in a group or dyad situation) is required

for toddlers' play behaviours to change. The project reported on in this thesis was

conducted in early-childhood settings where the program facilitates younger children

regularly engaging in group mixed-age free-play sessions. In addition the children

spent part of each day in free play with their same-aged peers.

Secondly in the dyadic play situation the researcher determined that a request for

the older preschool-aged peer to 'help' the toddler participant to play with the materials

provided was warranted.

2.12.3 Length and Frequency of Observations

O n reviewing the literature it seems imperative that multiple observations of the

children in each condition are warranted. This statement arises from two main

premises: first, play takes time and requires opportunity to develop, and secondly, a

child's play m a y vary considerably from day to day depending upon a number of factors

including 'within-child' characteristics (health, temperament, level of stress). Finally,

the toys and peers available to support children's play evoke a variety of play

behaviours.

In the current project it is planned to observe the toddler participants on three

occasions in each of three conditions. A s each observation is planned to last for ten

minutes a total of 180 minutes, video-taped data will be collected for each participant.

84


2.12.4 Programming for Mixed-Age Groups in Early-Childhood Settings

Successful programming for mixed-age groups in early-childhood settings

requires the staff to plan for and provide a range of appropriate materials for children of

varying developmental status. Sundell's work (1994) suggested that one method of

programming, employed by teachers operating in Swedish sibling-group classes where

the difference in age between the youngest and oldest child could be as much as six

years, was to allocate children to sub-groups based on age. This practice counteracts the

philosophical ideals supporting mixed-age grouping but illustrates the difficulties faced

by staff dealing with groups of children of widely disparate ages.

In the current project children were in groups where ages spanned a maximum

of four years and were together in mixed-age groups for part of their day only.

Nevertheless programming required close attention to the provision of toys and play

materials to support the development of symbolic play - the dependent variable in this

study.

2.12.5 Age Range of Children in Mixed-Age Groups

The findings revealed in both Sundell's work with Johansson (Johansson &

Sundell, 1989; cited in Sundell, 1994) and in his 1994 review indicate that the practice

of grouping children together where there is a large age span between the youngest and

the oldest child may not be beneficial. Sibling grouping as practised in Sweden would

seem to be atypical. In other countries, as in Sydney, Australia (the setting for the

current project) the ages of children in mixed-age groups rarely span more than three

years with a m a x i m u m of four years. The current study provides the ideal opportunity

to examine outcomes resulting from observing mixed-age groups reflecting a more

restricted age range. Findings from this current study are therefore likely to be more

generally applicable in the field of early-childhood than those of the Swedish studies.

2.12.6 Selection of Older-Peer Partner

A recurrent theme in the literature reviewed is the benefit of the younger child's

familiarity (perhaps indicating some existing relationship) with the older peer involved

in mixed-age dyads. In this present project the preschool-aged peer partnering the

participant in dyad play has been identified by staff as a familiar play partner of the

participant in group mixed-age play. Hence the older peer can be considered as 'self-

chosen' by the participant.

85


2.12.7 Caregiver or Teacher Influence

The current project poses the question whether staff training and qualifications

are influential on the development of toddlers' symbolic play in L D C C . It is predicted

that caregivers trained in Early-Childhood Education would be more supportive of

children's developing symbolic play, and actively seek to facilitate this play through the

provision of developmentally appropriate play materials. A questionnaire was designed

for the L D C C C staff to complete which first contained a number of questions designed

to assess staff attitudes toward play and secondly enabled the recognition of their

training and level of qualifications in Early-Childhood Education.

Rothstein-Fisch and Howes (1988) found that peer play (in Family Day Care)

was more complex when the caregiver was not engaged with the children but was

carrying out other duties. This is in contrast to the findings of Howes and Smith (1995)

that, in L D C C C s , positive social interaction with teachers was one predictor in the

competency variability in children's cognitive activity.

It is possible that, as the Family Day Care environment is more akin to a home

environment, the Family Day carer's activities are more aligned to those of maternal

carers in the home. Again, the consideration of a range of research directs the writer to

question the findings of Rothstein-Fisch and Howes (1988). A number of studies,

observing mothers and young children together in the home, have identified components

of the mothers' participation in and contribution to the play situation as supportive and

facilitative toward the development of young children's play.

While Family Day Care carers undergo training in the care of children, an in-

depth knowledge of child development together with an understanding of the

development of symbolic play is unlikely to be included. It m a y be that the caregiver's

interactions in Family-Day-Care conditions were not in tune with the children's needs

and so redirection or other intervention may have occurred.

2.12.8 Older-Peer Communicative Behaviour

The findings of the studies of older peers engaged with younger peers in either a

mixed-age group of mixed-age dyad play suggest that the older peers engage in

behaviours likely to support the younger children's play activity. Such support may be

verbal or behavioural. It therefore seems useful to identify and note the older peer's

communication in terms of the amount of information given and support for the younger

child's play during the mixed-age and dyad conditions.

86


2.12.9 Toddler-Participant Imitative Behaviour

In the present study it is likely, that in mixed-age play and in dyad play

(Conditions T w o and Three), the toddler will imitate the play activity of the older peer.

Such a prediction is in keeping with the findings of researchers whose work is explored

in this chapter in part 2.4. N o guidance is to be given for play to the older peer. (It is

thought that any guidance or suggestions made by the current researcher would detract

from the symbolic play likely to be elicited from the older peer by the play materials

and consequently might contaminate the data.) The toddlers' verbal or gestural

imitation of the older peer will be noted.

A broad range of the literature most relevant to the current investigation has

been reviewed in this chapter. This process had a two-fold purpose: first, implications

for the methodology arose from this literature review and were described briefly above.

(Each is discussed in more detail in Chapter Five.) Secondly the literature review

illustrated the range of theoretical perspectives it is essential to consider when

implementing a project such as the current investigation. Chapter Three explores the

theoretical bases underlying the present study.

87

Chapter 3: Theoretical Background

CHAPTER 3: THEORETICAL BACKGROUND 3.1 INTRODUCTION

The previous chapter contains a review of the literature relevant to the current

enquiry, which aims to identify some the influences on the development of symbolic

play in toddlers aged 18 months to 30 months. The literature appraised confirmed first,

the 'usefulness' of the major hypothesis and the three additional research questions

(presented in Chapter One, 1.8) and secondly the theoretical bases identified as

underlying this enquiry. In addition this review process enabled the resolution of many

methodological issues - these are described in Chapter Five.

3.2 THE AIM OF THE PRESENT STUDY

The current project, as mentioned above, is concerned with the examination of

some of the influences on the development of symbolic play in toddlers and is largely

grounded in the theory of two major cognitive theorists - Piaget and Vygotsky.

Consideration will also be given to the work of Rogoff who, in extending Vygotsky's

ideas, has formulated the notion of "guided participation" (1990, p. viii). The concept of

guided participation is highly pertinent to the study as participants are observed

interacting in free play with their older preschool-aged peers in two conditions

(Condition T w o , mixed-age play and Condition Three, dyad play).

The first mixed-age situation (Condition T w o ) comprises outdoor free play

where children ranging from approximately 18 months of age to a little more than five

years old share the same play environment. The second mixed-age situation (Condition

Three, dyad play) consists of two children - a toddler participant and his or her 'self-

chosen' familiar older peer - in an area aside from, but within, the children's main

activity area. Play materials to support symbolic play are provided for dyad play and

the older peer is requested to assist the younger child in play.

It is predicted that the two conditions described above will permit the toddler to

engage with preschool-aged peers in 'guided participation'. Rogoff, Mistry, Goncu, and

Mosier define 'guided participation' as "the process and system of involvement of

individuals with others, as they communicate and engage in shared endeavours" (1993,

p. 6). It is envisaged that both the mixed-age play conditions present contexts that

support the process of guided participation

88


A brief recapitulation of neo-Piagetian theory will also be included with specific

reference to the work of Case (1985, 1991), Halford (1993), Siegler (1991) and Fischer

and Farrar (1987).

3.3 THEORISTS UPON WHOSE WORK THE STUDY IS BASED

The initial impetus for the study of the development of symbolic or pretend play,

however, m a y be viewed as stemming from Freud's psychoanalytical theory (Winnicott,

1971); consequently, the psychoanalytical perspective, where relevant, will be

considered. Reference to Erikson's psychosocial theory is also pertinent (1963).

Erikson suggested that children resolved 'stage-like crises' at particular times in their

development. The importance of symbolic play can be seen as it occurs in the second of

Erikson's suggested psychosocial crises - autonomy versus shame and doubt - whereby

toddler-aged children strive for independence (1963).

Bandura's (1977, 1986) theory of social learning, later modified to a theory of

cognitive social-learning, is also appraised for its relevance to the child's use of

imitation in the acquisition of new skills. Bandura extols the value and efficacy of a

known or familiar and admired 'other' demonstrating or modelling behaviours as those

most likely to be imitated by the less skilled.

To a lesser degree theories of ethology and ecology will be considered.

Ethological theory and adaptation will be examined in the light of Burner's (1972, 1973)

discussion of the relationship of the nature of play and its role and purpose in

immaturity. Bjorklund (1997) also refers to this.

Reference to the work of Bronfenbrenner (1979), an ecological theorist, who

first suggested the notion of ecological systems theory, is also seen as applicable to this

project. H e postulates that children's development is impacted upon by a complex

network of relationships that surround children at a number of levels ranging from those

in children's immediate environment to those at institutional level. Where relevant,

connections to ecological theory will be demonstrated. Table 3.1 lists the theories to be

explored together with the proponents of such theory and the relevant links to the

current project.

89


Table 3.1

Theoretical Approaches and their Relevance to the Current Study

Theory Proponent(s) Relevance to current study

Cognitive constructivist

theory

Piaget(1896-1980) Deferred imitation,

transformations, internalised

schemas and representation of

those schemas.

Imitation viewed as practice

play.

McCune-Nicolich (1977, 1981) Assessment of children's

symbolic maturity, importance of

language in decontextualisation',

notion of 'multischeme

combinations'

neo-Piagetian theory Case (1991) Importance of cultural context

Working memory

Halford (1993) Importance of environment short

term memory

Fischer and Farrar (1987) Generalisation

Performance levels

Cognitive socio-cultural

theory

Vygotsky (1896-1934) Socio-cultural context, 'zone of

proximal development', rules of

society

Rogoff (1990, 1993) 'Guided participation'

'The more-skilled other'

Cognitive social-learning

theory

Bandura (1977, 1986) Observational learning

Imitation of others' behaviours

90


Theory Proponent(s) Relevance to current study

Psychoanalytic theory Peller (1954, 1978) Symbolic play allows child to

express and discharge motions,

and feelings, requires 'low-level'

anxiety

Winnicott (1965, 1971) Symbolic play develops as

'separateness' occurs between

child and caregiver

Psychosocial theory Erikson (1902-1994) Symbolic play important

component of second

psychosocial crisis (autonomy

vs. shame and doubt)

Ethological theory Bruner (1971, 1974) Symbolic play allows practice of

life skills in secure environment,

tutor proneness.

Ecological theory Bronfenbrenner (1979, 1993) Child operating in microsystem,

importance of "proximal

processes" (1993, p. 8)

Bjorklund (1997) The adaptive role of

developmental immaturity

91


3.4 THE COGNITIVE FUNCTION OF SYMBOLIC PLAY

As stated initially the predominant component of this discussion, examining play

from a cognitive theoretical perspective, will be with reference to Piaget (1962) and

Vygotsky (1962) - two major theorists in the area of child development. Because they

have refined some of Piaget's theory and are more contemporary, neo-Piagetian

theorists and the components of their theories relevant to the present study are explored

in 3.3.2 of this chapter following a detailed analysis of the Piagetian perspective of

symbolic play (3.3.1).

Since it is so comprehensive and widely regarded, the Piagetian theory of

cognitive development will be closely examined with particular attention paid to both

his description of symbolic play and the ideas of assimilation and accommodation.

Although play was perhaps not a major interest of Piaget, his theory of cognitive

development is nevertheless particularly relevant in terms of both his description of

symbolic play and the processes of assimilation and accommodation.

It is maybe useful here to be reminded that Piaget's theory of cognitive

development comprises four main stages or periods encompassing a number of

substages. Piaget believed that these four stages were sequential and invariant. The

participants in this current project bridge two of the stages of intellectual development

identified by Piaget. The younger participants (those aged approximately 18 to 24

months) are likely to be in the sixth substage of the sensori-motor period - the first

Piagetian stage of cognitive development.

The cognitive behaviour of children in the sixth substage of the sensori-motor

period - intellectual activity described by Shaffer, as the "invention of new means

through mental combinations" (1989, p. 314) - is characterised by two abilities essential

for the successful engagement in symbolic play. The first of these is the capacity to

engage in systematic deferred imitation - accomplished Piaget suggests "through the

interiorisation of accommodation" (1962, p. 84) - and the second, the aptitude to

construct mental symbols.

The older participants (those aged approximately 24 months and over) would be

expected to be in the preconceptual substage of the second stage of cognitive

development described by Piaget, the preoperational stage. (While the preconceptual

stage spans the years from approximately two to four the oldest participant in this

present study is 31 months old.) Children in the preconceptual stage of cognitive

development are likely to be capable of operating at a symbolic level by incorporating

92


words or objects to represent supported by the use of language. These abilities allow

them to use words and objects in conjunction with activity to represent components of

their environment. However, as the name preconceptual suggests, ideas and concepts

may remain largely undeveloped and lacking sophistication.

Piaget recognised play as a means of facilitating intellectual growth and

development. H e viewed play as a way for children to practise and consolidate

previously learned skills and so adapt to the environment both physically and

intellectually. However, Piaget did not identify play as synonymous with learning but

suggested that when engaged in play activities children were able to practise their skills

and competencies in an informal and light-hearted manner. This view is not dissimilar

from that held by Bruner (1972), who describes play as giving young children the

opportunity to practise life skills in a 'relaxed' way because as yet, their survival is not

dependent upon those skills. H e suggests that in play children are able to try

"combinations of behaviour that would, under functional pressure, never be tried",

(1972, p. 693).

Symbolic play, Piaget (1962) asserted, illustrated children's capacity to use

insight and mentally process problems. Play, he argued, was the unbalanced (but

normal) state experienced by children when accommodation remains unresolved and

assimilation predominates. Nevertheless, the Piagetian analysis of play identifies some

element of adaptive function that to some degree is in keeping with the findings of those

theorists adhering to an ethological model, such as Bruner w h o recognised the role of

play in the adaptive behaviour of the young. Piaget's contribution is analysed in the

next Section of this chapter (3.4.1).

Vygotsky (1962) was a social-constructivist concerned with cognitive

development from a socio-cultural perspective. His philosophical background was not

too dissimilar from that of Piaget. Vygotsky opposed the notion of competency as

prerequisite to performance with the claim that performance can indeed precede

competence. That he viewed play as important is illustrated metaphorically as "as in the

focus of a magnifying glass, play contains all the developmental tendencies in

condensed form and is itself a major source of development" (1978, p. 102).

Vygotsky identified the importance of the social context in which a child's

experience takes place together with the value of collaboration between the child and a

more experienced other in the acquisition of new knowledge, skills or learning. This

idea can be recognised in what Vygotsky termed the 'zone of proximal development'

93


referring to the difference between a child's ability, for example, to problem-solve

independently compared to his or her ability when supported or aided by a more

competent other. However, Vygotsky considered the potential a child could reach with

adult or peer support was limited by the child's maturational processes. Operating in

conjunction with the child, the collaborators, the 'more-skilled others', provide a

socially-mediated context. Vygotsky further proposed that the development process

lags behind the learning process and that good teaching proceeds just a little ahead of

development and makes use of the child's functions that are maturing. The Vygotskian

approach m a y also be likened to the idea that learning is the "buds and the flowers" of

development and not the "fruits" as in the Piagetian model (Tudge & Winterhoff, 1993,

p. 67). Vygotsky's contribution is considered further in Section 2.3.3 of this chapter.

In summary, the major difference between the two theorists, Piaget and

Vygotsky, in terms of the social nature of the child, is that, while Piaget argues that the

child becomes progressively more social as cognitive development becomes more

sophisticated, Vygotsky describes the child as an inherently social being from the

beginning of life. Implied in the latter premise is the importance of adults and peers, as

facilitators and collaborators, in the learning process. The different approaches toward

play taken by the two theorists are summarised succinctly by Ebbeck w h o suggests that

Piaget would define play as "exploration" and Vygotsky as "an outcome of social

support" (1996, p. 27).

In conclusion it is apparent when considering play from either a Piagetian or a

Vygotskian perspective that a number of important variables pertaining to children's

early experiences and cultural contexts and the effect of those upon play are worthy of

discussion. Variables include: parenting style - a variable likely to be dependent upon

the cultural context of the family, the position of the child within the family and the

socio-economic status of the family unit. The nature of attachment (related to parenting

style) is worthy of attention in that an infant's play behaviour is synonymous with his

exploration of the environment. Successful exploration is more probable if the infant is

securely attached to the caregiver (Slade, 1987b; Meins, Fernyhough, Russell, & Clark-

Carter, 1998).

94


3.4.1 A n Analysis of the Piagetian View of Play

Piaget (1962) considered play as representative of the unbalanced state

experienced by children when they were in the process of interpreting new experiences

and modifying existing schemata - in Piagetian terms, accommodating and assimilating.

Although Piaget's cognitive theory identifies stages of cognitive development, included

in the theory is his perspective on the cognitive development of play. Children, he

suggested (Piaget, 1962), experience three different types of play between birth and

twelve years of age. It is important to realise that these categories of play are not

discrete and can be seen to overlap in the different developmental stages reflected in the

cognitive and physical domains.

In order to appreciate and comprehend the sequence of play development

involved in the production of symbolic play, it seems useful to examine first the

category of practice play. This type of play occurs before symbolic play emerges with

the final stage of practice play, "intentional combinations" appearing to serve as a

precursor to the development of symbolic play (1962, p. 116).

Infants and toddlers, in the first two years of life, according to Piaget, are

involved primarily in practice play. The content and activities manifested in practice

play are closely aligned with skills developed during the sensori-motor stage of

cognitive development. The use of the senses, combined with physical activity, allow

young children to interact and explore first their o w n bodies, and later components of

the environment including play materials.

Piaget (1962) considered that practice play embraced three sub-categories or

types of play depending upon how much of the child's intention and integration of

acquired skills and knowledge was evident. Early practice play involving young

children's discovery of the effect of physical activity, for example, reaching or

perfecting some skill often involving their own bodies, was referred to by Piaget as

"mere practice" (1962, p. 114) - the first type of practice play to develop. The first

early examples of 'mere practice' play coincide with the stage of cognitive development

in the sensori-motor period of primary circular reactions - whereby the infant's chance

discovery that his various actions or responses are both appeasing and able to be

repeated. 'Mere practice' play, while tending to be the play most often engaged in for

the first eighteen months of life, can, of course, recur later in life when specific skill

development is required.

95


The degree of playfulness integral to 'mere practice' play is difficult to define,

and determination m a y depend upon other existing conditions. For example, if

repetition of an activity is undertaken voluntarily to gain mastery or to maintain a

function, then Piaget would judge that play is in existence. In this case the control is

internal, exercised by the player. However, if an external control is exerted, then the

resultant 'mere practice' play is thought to be less playful (Smith & Vollstedt, 1985).

'Mere practice' play m a y also involve repetition of skills in which the

participant is already proficient. Such activity may then result in maintaining a

function, which again would be recognised by Piaget as play. Piaget (1962) considered

that 'mere practice' play does not always entail physical activity, and concluded that

mental exercise, that is, activity of the mind (schemas), can also constitute 'mere

practice' play.

During this initial stage of practice play, Piaget suggests that the child may

inadvertently combine familiar knowledge or schemas with new behaviours. Piaget

named these incidents "fortuitous combinations" (1962, p. 115) and identified them as

the second category of play in the practice-play stage. Such combinations are aligned

with responses in cognitive development that Piaget named 'secondary circular

reactions' - also in the sensori-motor period - in that the child is able to enjoy a

satisfying experience by acting upon objects other than his or her o w n body. After the

initial discovery, these acts can be repeated at will; nevertheless, such acts were not

considered by Piaget to be the result of intent as the first event was accidental.

'Fortuitous combinations' are sometimes manifested in what can be termed specifically

playful behaviour, which m a y contain an element of surprise in reaction to the novel

event. Again, there would appear to be no design or anticipation, but the child is likely

to give some behavioural or even verbal indication that fun or playfulness is a

component of the incident. As with 'mere practice' play, 'fortuitous combinations',

while initially a feature of the very young child's play, m a y occur at other times in life.

The final type of play, identified by Piaget as practice play, is that which is

termed "intentional combinations" (1962, p. 116). Encompassed in this type of play is a

degree of both anticipation and planning or intention. Such play involves

experimentation as children use trial and error to produce interesting and pleasing

results. In terms of cognitive development this type of play parallels children's

progression into tertiary circular reactions - a substage in Piaget's sensory-motor period

when they are able to planfully act upon objects in their environment. Play using

96


'intentional combinations' illustrates children's early attempts to employ insight and

enables them to be quite systematic in their exploratory actions. Signs of intent and

planning are evident together with the ability to imitate behavioural sequences at a later

date. Piaget (1962) postulated that in this stage of tertiary circular reactions children

remain unaware of any impact of their actions on the environment as they are firmly

focused on the acts of combining and repetition.

The activities in these categories of practice play described above can be seen as

being instrumental in equipping children with the necessary skills required for them to

engage in symbolic play. Pretend or symbolic play signals the onset of children's use of

objects, behaviours such as gestures and postures, as symbols or representations of

cognitive activity removed from the reality of the moment (Piaget, 1962).

To comprehend Piaget's perspective on symbolic play it is useful to think of

such play as allowing players to exhibit their knowledge of "knowing-how" and

"knowing-that" before they could, in real life, actually carry out the procedure they are

depicting through play (Stambak & Sinclair, 1993, p. xi). Piaget suggests that symbolic

play also allows the observer to acquire an understanding of the child's emotional state

through the verbal or non-verbal behaviour expressed in the symbolic-play episode. In

keeping with psychoanalytic theory in respect of symbolic play, Piaget postulated that it

not only "enables the child to relive his past experiences" but also "has as its sole aim

satisfaction of the ego" (1962, p. 167). Also in the psychoanalytic vein symbolic play

assists the player to discharge emotion by offering an opportunity to resolve personal

conflict (Stambak & Sinclair, 1993).

Let us briefly turn to some implications of Piagetian theory to the present study,

which requires the observation of toddlers engaged in play in three conditions. First,

toddlers will be observed in Condition One, free play with their same-age peers;

secondly in Condition T w o , free play with older preschool-aged peers, and finally, in

Condition Three, with an older peer in dyad free play. Props for the first two conditions

will be those supplied by the staff in the L D C C C staff as part of their usual

programming procedure while those for Condition Three will be supplied by the current

researcher.

It is expected that the younger child in Condition Three (dyad play) will engage

in a more sophisticated type of practice play than would be possible for the child alone

(a situation clearly illustrating Vygotsky's [1978] theory of Z P D ) . Piaget (1962)

viewed practice play as having several levels of development - the final most

97


sophisticated level containing an intentional planned component. From a Piagetian

perspective it m a y be that dyad play supports what Stambak and Sinclair describe as

children engaging in "reciprocal propositions that can be coordinated" (1993, p. xiii).

The older child, in offering suggestions either verbally or through gesture, may elicit a

reciprocal action from the younger child. These suggestions m a y take the form of

propositions (as suggested by Stambak and Sinclair, 1993) and the subsequent

coordination of activity by the two players m a y be an inadvertent outcome of Condition

Three. In the present study, dyad play will take place in a less distracting environment

than that experienced by children in group free play, whether with same-age or mixed-

age peers.

Instances of imitation in by the toddler participant in dyad play will also be

noted. Theorists vary in their explanation of imitation or imitative behaviour. For the

purpose of the current investigation the theoretical perspectives of Piaget, Vygotsky and

Bandura are relevant to describing the imitative process. In Piagetian terms the

imitation likely to be exhibited by a toddler-aged child (18 to 30 months of age) as "a

continuation of accommodation for its own sake" (Piaget, 1962, p. 87). Piaget appears

to very clearly separate imitation from play when he describes play as "essentially

assimilation or the primacy of assimilation over accommodation" (1962, p. 87). These

statements convey the notion that Piaget perceived play as the exploration and trial of

novel unpractised activity, whereas imitation falls under the category of practice

behaviour. Piaget also views imitation as a developmental feature of the toddler period,

a refinement of the initial imitation activity, which emerges during the fourth substage

(coordination of secondary schemata) in the sensori-motor period. Imitation can also be

described in terms of social-learning theory, which is dealt with later in this chapter in

3.5.

Finally, reference to the work of McCune-Nicolich (1977, 1981) is essential

when investigating symbolic play development from a Piagetian perspective. Her

studies resulted in an analysis of the development of symbolic play, over the period of a

year, of children aged 14 to 19 months. This analysis enabled McCune-Nicolich to

assess the level of children's symbolic maturity. Later symbolic play, the development

of symbolic functioning, and in particular the reciprocal links between symbolic play

and early language, were of particular interest to McCune-Nicolich. She postulates that

this later period of symbolic play, labelled 'decontextualisation', requires prior planning

and the mental construction of action before engagement. Therefore such play is

98


independent of available equipment, context or materials. (In Westby's [1991] Scale for

Assessing Children's Pretend Play, which is to be used in the present study,

'decontextualisation' coincides with Dimension One of symbolic-play behaviour.)

McCune-Nicolich (1981) also proposed that this advanced level of pretence

requires the child to integrate several schemes - she described this process as

'multischeme combinations' - citing the child's developing language as reflective of this

transformation in symbolic play. Her understanding of the interdependence between the

symbolic-play behaviour of 'decontextualistion' and language development is integral

to any investigation of emerging symbolic activity in toddler-aged children.

3.4.1.1 Parallels Between Piaget's and Westby's View of Symbolic Play

As foreshadowed in Chapter One (1.4.1), Westby's (1991) Scale for Assessing

Children's Pretend Play which effectively allows the coding of identification of changes

in toddlers' pre-symbolic play and symbolic play will be used in the present enquiry to

assess the level of the participants' symbolic play. (This scale is presented in Appendix

Ai and discussed in more detail in Chapter Four, 4.5.) It seems useful to now explore

the parallels between the components of symbolic play according to Piaget and those

dimensions of symbolic play identified by Westby (1991).

'Decontextualisation' (Dimension One)

The first dimension of symbolic play according to Westby (1991) is

'decontextualisation' or 'object substitution'. In terms of Piagetian theory the toddlers

in this study, aged from 17 to 30 months, straddle two stages of cognitive development.

Those aged from 17 to 24 months (about half the participants) are likely to be in what

Piaget named the sensori-motor stage of development - the first stage of cognitive

development. The remainder of the participants, aged approximately 25 to 31 months,

are likely to be in the stage identified by Piaget as the preconceptual stage - the first

component of the preoperational stage; Piaget's second stage in cognitive development

(1962). During both the latter part of the sensori-motor stage and the duration of the

preconceptual period of the second stage of cognitive development, Piaget proposed

that, while children are capable of thinking at a symbolic level, to manifest actions

linked to this thought they require the support of concrete materials (1962).

99


'Thematic Content' (Dimension T w o )

The second of Westby's (1991) dimensions of symbolic play is 'thematic

content' which refers to the themes or scripts evident in children's play. Initially such

themes are concrete progressing over time to those that are more abstract. Piaget (1962)

viewed symbolic play as having three types - the second of which, whereby children

imitate the behaviours of others in their environment, m a y be likened to Westby's

notion of Dimension T w o symbolic behaviour - 'thematic content'. Importantly, such

imitative behaviour utilises the skill of deferred imitation - a 'signifier' in Piagetian

terms indicting the emergence of the 'semiotic function'. (Deferred imitation in

Piagetian terms emerges in the tertiary circular reactions stage of the sensori-motor

period - when the child is approximately 12 to 18 months old.) Deferred imitative

behaviour at this stage does not involve the 'taking on of another's role or persona' (in

Westby's terms the Dimension-Four symbolic behaviour of 'self-other relationships') -

it is purely the behaviour of the 'other' that the child copies. C o m m o n examples of

'thematic content' gleaned by children from another's repertoire are the actions of a

parent - again likely to be those described as 'everyday' or familiar routines or actions.

McCune-Nicolich, confirms Piaget's process of symbolic development, in her

work published in (1977). Using a more detailed scale of levels and transition criteria

(though maintaining a strong relationship to the symbolic stages suggested by Piaget),

McCune-Nicolich suggested that a child's 'sourcing of a scheme" from another's

activity confirmed the "separation of the symbols from the child's sensori-motor

actions" (McCune-Nicolich, 1977, p. 95).

'Organisation of themes' (Dimension Three)

As mentioned above, Piaget viewed symbolic play as having three components

only. Consequently, the third dimension suggested by Westby (1991) 'organisation of

themes' does not parallel a specific Piagetian concept in symbolic play.

'Self-other relationships' (Dimension Four)

The fourth dimension of symbolic play as described by Westby (1991) ('self-

other relationships') is closely affiliated with the Piagetian notion of 'decentration', the

ability to separate one's self from the environment (1962). Successful decentration,

Piaget declared, "results in equilibrium between assimilation and accommodation, an

equilibrium which of necessity tends towards a reversible structure" (1962, p. 243).

100


Piaget went on to predict that decentration, the process of decentring from the self, is

finalised when aged approximately 18 to 24 months - the period in Piagetian cognitive

theory, which marks the sixth and final stage of the sensori-motor period. The child's

acquisition of decentration allows him to act upon objects or others in his environment.

Before this progression in development the child's actions in terms of symbolic play are

likely to be directed toward him and termed, as Westby (1991) suggests, 'auto-

symbolic' or 'self-representational' acts.

As discussed earlier in this chapter, in contrast to Piaget's predictions, McCune-

Nicolich (1981) suggests that young children achieve decentration earlier than 18

months of age. Westby (1991), who appears to follow Piagetian theory, suggests that

children aged 17 to 19 months certainly engage in self-directed play. However, as

reflected in Westby's Scale for Assessing Children's Pretend Play (1991), the

expectation is that children have, by 19 to 20 months, progressed to including objects or

other people as passive recipients of their symbolic activity.

3.4.2 An Analysis of the neo-Piagetian Perspective

As mentioned earlier the current study is based mainly on the work of the two

perhaps most prominent cognitive theorists - Piaget and Vygotsky. However to

maintain and support a contemporary approach it is essential to consider elements of

neo-Piagetian theory. To that end it is necessary to discuss briefly the components of

several representatives of neo-Piagetian thought.

Case (1985) welds together the cognitive theory of Piaget with the contemporary

information-processing approach. H e suggests that there are four main stages of

development each of which is concerned with the process as well as the structures of

thinking (1991, p. 48). Case contends that changes in a child's development are brought

about by changes in the mind's general capabilities and suggests that once a mental

capacity is achieved - far from being 'domain-specific' - that capacity or growth can be

successfully applied to other domains (1991, p. 35). In terms of symbolic-play

development, Case's description of short-term memory use is of interest. H e postulates

that the acquisition of a new skill initially requires considerable support from short-term

memory. However, in subsequent practice of this skill or 'automisation' the

dependency on short-term memory is lessened, thus relinquishing space in that memory

for the acquisition of other cognitive skills or learning (1985). (The notion of

'automisation' coincides with what Case describes as 'postulate 7' of a model of stage

101


transition - namely that, "the growth of short-term storage space within each stage is

brought about by an increase in operational efficiency," 1985, p. 414.)

Case (1991) is also concerned with the cultural context of development and

suggests that the challenges or problems that children encounter are 'culturally bound',

and are increasingly so the more abstract problems become. Successful resolution of a

problem requires the child to learn experientially, in cultural contexts, and to learn from

others in a given context w h o model appropriate and successful ways of dealing with

challenges.

The work of Halford (1993), another neo-Piagetian, is also pertinent to this

project in two main ways. Halford proposes that both the environment and innate

factors are influential on a child's understanding and cognitive development. He, like

Case, incorporates within his theory aspects of information-processing theory by

declaring "understanding plays a crucial role in cognitive development but it needs to be

considered in relation to other basic processes such as learning, memory and

development of capacity" (1993, p. 1). Halford recognises the importance of others in

the child's learning process and, while agreeing "cognitive development does not

proceed solely by spontaneous exploration but is aided by social influences and

insight", proposes also that "children's knowledge is not a gift from his or her mentors"

(1993, p. 14-15). In keeping with Piagetian theory, Halford views children as active

participants in constructing their own learning and emphasises that it is through their

activity, not instruction alone, that concepts are acquired (1993, p. 275).

In relation to a child's acquisition of symbolic play skills, two of Halford's

notions seem particularly relevant. First he considers that, for social input to be useful

to a child it must be related to the child's own representations (1993, p. 275) and

secondly he distinguishes 'declarative' knowledge - the "knowing that'", from

'procedural' knowledge - the "knowing how" (1993, p. 16). Halford surmises that the

former - the 'knowing that' knowledge - is held in the long-term memory, while the

latter - the 'knowing how' knowledge - relies upon short-term memory activity. This

idea perhaps parallels that of Case, in that once the 'declarative' knowledge is acquired

and resides in the long-term memory, the short-term memory is freed-up to engage in

the acquisition of further 'procedural' knowledge.

Fischer and Farrar (1987) subscribe to a skill-theory framework approach - one

that combines elements of both behaviourist and information-processing theory - to

explain children's development and learning (p. 139). Fischer and Farrar (1987) used

102


the symbolic-play process (whereby activities experienced or observed by children

earlier in other contexts are represented in play) as an example to illustrate the authors'

notion of generalisation - a component of the skill-theory framework approach.

Additionally, of interest to the present researcher is the notion proposed by Fischer and

Farrar of a child's differing levels of performance: - an optimal level which occurs in

familiar, well-supported environments; and a junctional (lower) level which occurs in

what Fischer and Farrar term "a given sub-optimal context" (1987, p. 159). Both

generalisation and performance levels are useful when explaining children's early

symbolic play activity and development.

In summary, the neo-Piagetians have contributed to the knowledge of children's

cognitive development in important ways. B y supplementing some Piagetian principles

they have perhaps 'wedded socio-cultural theory to information-processing theory'

while from Fischer and Farrar's perspective, allowing behaviourism to be 'in

attendance'. Their work alerts researchers to the importance of context and "transition

mechanisms" - those points in development or events in experience that allow an

increase in children's cognition (Demetriou, 1988, p. viii).

3.4.3 An Analysis of the Vygotskian View of Play

Vygotsky (1987), a socio-cultural theorist, while in agreement with Piaget in

terms of the child's being an active learner and constructor of knowledge, emphasised

the importance of the social context in which this exploration and learning takes place.

This recognition of social context contributing to the learning experience is related to

Vygotsky's belief that generally a child's 'performance' preceded his or her

competence. In support of these two premises, Vygotsky viewed the more-able 'other'

- peer or adult - as instrumental in supporting the child in constructing new knowledge

and new cognitive processes. These collaborators he considered as taking the role of

'co-constructionists'. Vygotsky construed the concept of the 'zone of proximal

development' - a key notion in his approach. This, he suggests, refers to the difference

between the child's unaided competency in some skill, knowledge or area of learning

and the level of skill, knowledge or learning which the child could attain with an older

(or more-accomplished) peer or adult support. However, Vygotsky recognised that both

heredity and the maturation process limited the potential achievable by the child. H e

103


further postulated that the developmental process lags behind the learning process and

that "instruction is only useful when it moves ahead of development" (1987, p. 212).

The social nature of children perhaps highlights the greatest area of dispute

between Piaget and Vygotsky. Piaget proposed that, as children's cognitive

development becomes more sophisticated they, in turn, become progressively more

social. Members of the Vygotskian school would describe children as a social beings

from the beginning of life - hence the importance of social interaction with adults and

peers in the learning process. This perspective contributes to the differing stances on

play taken by the two theorists. Vygotsky contends that children involved in play are

behaving above their age and displaying behaviour beyond that which would normally

be expected from them - "in play it is as if he were a head taller than himself"

(Vygotsky, 1978, p. 102). H e further postulates that "the child moves forward

essentially through play activity" and that play can "be termed a leading activity which

determines the child's development" (Vygotsky, 1966, p. 16). Importantly, the

Vygotskian approach, reflecting as it does social and collaborative components, allows

children, through play, to both operate in and acquire the attributes and features of their

cultural context.

Imitation, a developmental process, features in both Piagetian and Vygotskian

theory, though possibly the latter placed more emphasis on the potential of the imitative

process. Vygotsky believed that, "imitation is the source of instruction's influence on

development and that instruction was only possible where there is potential for

imitation" (1987, p. 210-211). Again, while realising the constraints placed upon the

individual by heredity and maturation, Vygotsky surmised that imitation was the avenue

through which children were able to cognitively act upon the influences encountered in

the environment. Play was the medium that released children from the restraints of

reality and allowed them to engage in imaginary situations. Cognitive growth is

promoted as children's behaviours or actions within a play episode take on symbolic

meaning. In keeping with Piaget, Vygotsky viewed symbolic play as requiring children

to separate meaning from actions and objects - promoting representational thought

processes based on these meanings.

It maybe useful to identify and consider what would seem to be opposing

ideological perspectives evident in the theories of Piaget and Vygotsky. Piaget's notion

of the child as a 'lone venturer' is closely related to the goals of a Western-orientated

society where individual achievement is viewed as desirable and most worthy of

104


acclaim. Bruner noted that Vygotsky was "a pioneer of Marxist psychology" (1984, p.

93) and so was an advocate of the Eastern-European approach of more collective

responsibility towards a child's development.

3.4.3.1 Parallels Between Vygotsky's and Westby's View of Symbolic Play

It will be recalled that Westby's (1991) Scale for Assessing Children's Pretend

Play will be used in the present project to assess changes in children's symbolic play.

This scale allows the coding of symbolic play in four dimensions -


relationships'. It seems pertinent to now identify the components of cognitive

development according to Vygotsky that correspond or are similar to Westby's (1991)

dimensions of symbolic play.

'Decontextualisation' (Dimension One)

W h e n considering the first dimension of symbolic play 'decontextualisation',

Vygotskian theory closely parallels that of Piaget. Vygotsky recognised that a lessening

in dependence on life-like props for supporting symbolic play is closely linked with

children's cognitive development. H e postulated that, for young children, to separate

words (or thinking) and objects was a difficult task - accomplished gradually and over

time (1966). Hence toddler-aged children are less capable than their older peers of

sustaining symbolic thought - unless they are able to employ concrete objects to

provide a connection between thought and action (Vygotsky, 1966).

Westby (1991) contends that a reduction in this dependency on life-like props,

for successful symbolic play, is not discernible in young children until they are

approximately 36 months old - an age beyond the scope of this study. Hence, in terms

of the Westby Scale for Assessing Children's Pretend Play only two levels of

Dimension One exists - Level One (at 17 to 19 months of age) and Level Six (36

months of age) in the present study.

'Thematic Content' (Dimension Two)

The more sophisticated levels of 'thematic content' (Dimension T w o of

symbolic-play behaviour) begin with children replicating symbolically the actions of

others in their environment. Such events, when viewed from a Vygotskian socio-

cultural perspective suggests they evolve from children's close contact and interaction

105


with more-skilled others (parents, siblings, peers and caregivers). This involvement is

essential to guide and support young children's learning, development and acquisition

of appropriate societal rules, skills and behaviours.

'Organisation of themes' (Dimension Three)

'Organisation of themes' is the dimension of symbolic play for which the child

is initially likely to complete one single act of symbolic activity - later linking two or

more actions in a logical sequence. From a Vygotskian perspective the 'triggering' of a

sequence of actions may actually reflect the child's early activity in conforming to

social norms - the rules of society. Vygotsky would consider a child, who poured

imaginary liquid from a jug into a cup and then pretended to drink it, as exploring or

practising the rules of the situation represented in the play - "the imaginary situation

will always contain rules" (1966, p. 10).

'Self-other Relationships' (Dimension Four)

'Self-other relationships' (Dimension Four) of symbolic play (in Westby's

terms) is closely related to what Vygotsky viewed as 'action-meaning ratio'. He

postulated that the 'object-meaning ratio' and the 'action-meaning ratio' were the two

major components of play (1978, p. 100). As the player gives meaning to an object, so

he does to an action. 'Decentration' can be described in Vygotskian terms as the child's

detachment of meaning from actions. Vygotsky proposes that initially, in play, action

dominates meaning and that the child fails to fully understand many of the ensuing

actions - "the child is able to do more than he can understand" (1978, p. 100). In time

meaning 'overtakes' or dominates action - a process described by Vygotsky as evidence

of, "the development of will or the ability to make conscious choices" (1978, p. 101).

Vygotsky proposes that such a process represents "movement" of an abstract nature in a

context that is both "situational and concrete" (1978, p. 101).

3.5 SYMBOLIC PLAY AS A FUNCTION OF SOCIAL-LEARNING THEORY

Social-learning theory described, by perhaps its founder Bandura (1986), as

cognitive social-learning theory, is useful in this study for assisting in the identification

and understanding of imitative behaviour. In the current project, imitative behaviour is

likely to be exhibited by the younger peer in any of the three conditions: Condition One

106


-same-age play, Condition T w o -mixed-age play or Condition Three - dyad play. In a

cognitive social-learning context, imitation refers to observational learning whereby an

observer repeats a model's behaviour. Observation of a more-skilled other is one

method by which young children acquire new skills. Importantly, a higher frequency of

imitation occurs if the model is familiar and well liked by the observer (Bandura, 1971,

p. 17).

For successful observational or imitative learning to occur, Bandura (1971)

considers the following four processes as essential: attention, retention, motor

reproduction and reinforcement and motivational processes. These happen sequentially

and take place between the modelled event or behaviour and the imitative behaviour of

the learner. T o illustrate this process, for the toddler participants in this study to

successfully engage in imitation or observational learning they must first attend or

monitor the older peer's activity, use their memory skills to retain what has been

observed, be physically able to reproduce what was observed, and finally be motivated

to do so. T o complete the social-learning theory process, a reward or some form of

reinforcement is required following the imitative gesture or behaviour. T w o of the three

conditions in this current project - mixed-age play and dyad play — are play contexts

which have the potential to fulfil the requirements for the four processes proposed by

Bandura to occur (1971, 1977).

In the present study older peers are engaged in free play with the younger

toddler-aged participants on two occasions. The first, Condition T w o , is a mixed-age

free-play session, and the second, Condition Three, is dyad free play. It is in this last

condition (Condition Three) when the participants are partnered by their 'self-chosen'

and familiar older peers that the researcher expects to see the toddler engage in imitative

behaviour. A s mentioned earlier the model's attributes appear to contribute to the four

processes necessary for observational learning to occur. Bandura suggests that

"attention to models is also channeled by their interpersonal attraction" (1977, p. 24).

He continued "models w h o possess engaging qualities are sought out, while those

lacking pleasing characteristics are generally ignored or rejected" (1977, p. 24). It is

likely, as the older peers are the participants' 'self-chosen' familiar playmates, that they

meet those requirements suggested by Bandura.

As stated above to complete the social-learning theory process, a reward or some

form of reinforcement is required following the toddler participant's imitative gesture or

behaviour. Older peers acting as tutors or mentors are often skilled at fulfilling this

107


requirement. It m a y be that such behaviour is an example of what Bandura describes as

" . . . pleasing characteristics . . ." (1977, p. 24).

3.6 SYMBOLIC PLAY AS A FUNCTION OF PSYCHOANALYTIC AND

PSYCHO-SOCIAL THEORY

Play viewed from the psychoanalytic stance, has a substantive role in children's

emotional development, since through play children are able to express and discharge

intense feelings. Theorists taking a Freudian perspective believe that play m a y have a

cathartic effect on children in that, through the medium of play, children are permitted

to take control of and have an effect on the environment.

Peller (1954) suggests that such activity positively affects both the libido and

ego in terms of their interdependence. Her premise is that play is brought into being as

a result of ego activity, namely anxiety, in response to experiences encountered (reality)

together with issues coming to the fore as a result of id or superego activity. Peller

(1954) suggests that for play to occur the anxiety must be of low intensity and that play

would not be possible if a high level of anxiety were to persist. This notion of anxiety,

even a low-level anxiety, instigating play is in direct contrast to Bruner's (1973) notion

of play as a low-stress occupation and therefore one would assume, an anxiety-free

activity. O n e could speculate that the presence or otherwise of anxiety is what defines

the outcome of play as either a cathartic experience - as in the psychoanalytic tradition

- or a skill-acquisition exercise in keeping with Bruner's constructs.

The opportunity for children to be in control is especially important following an

unpleasant experience, and play allows them to release negative feelings or actions in a

socially acceptable manner. This experience of control is in sharp contrast to the child's

being a passive recipient, which tends to be the perceived reality for the child.

Psychoanalytic theory m a y also provide the basis for some categories of play therapy -

for example the opportunity for children to repeatedly act out part of an undesirable

episode allows them, in time, to deal with the negative feelings. The taking on of roles,

a symbolic activity, is synonymous with control and integral to the psychoanalytic

approach to play. Role taking is a component of symbolic play termed by Westby

(1991) as 'self-other relationships'. Initially, as symbolic play emerges, the play

activities are directed toward the self - the child is the agent. Later, as symbolic-play

skills develop and strengthen, the player is able to act upon first, objects (for example a

108


doll or teddy-bear), later 'others' (adults or other children) and finally the player can

adopt another's role.

In deliberating upon the development of symbolic play in collaboration with

others and taking a psychoanalytic viewpoint, a recognition of the intimacy of the other

child in the dyad and the content of the interactions between the two children occurs.

Winnicott (1965, 1971) suggests that symbolic play, the focus of this study, develops as

a result of the changing relationship between the young child and the mother or

caregiver. Werner and Kaplan (1963) noted the importance of the caregiver-child

interaction for symbolic play to develop, in that such play unfolds as a result of the

physical distance between the caregiver and child widening, so that subsequently play

becomes more shareable. Werner and Kaplan identified symbolic play and its inherent

skills, such as the ability to consider absent objects, people, events or situations and to

represent them, as one of the child's initial ways of expressing and experiencing

autonomy.

Finally, Peller, influenced by Freud's teachings, considered that "in play the

child learns to tolerate reality" (1978, p. 199) - the term 'tolerate' suggesting a 'coming

to grips with' or a 'practising of skills in readiness' for reality. If the participants'

capacity to engage in more-frequent symbolic play and at a more complex level than is

expected for their age was found in the present study then it could be concluded that

such 'toleration' m a y be expedited by free play with older peers.

Erikson, described as a psychosocial theorist, viewed play as a method by which

children practised or experimented with ways to master reality (1963). In keeping with

the Freudian notion of using play to resolve inner conflict or to become powerful in the

face of adversity, Erikson speaks of play as "the most natural self-healing measure

childhood affords us" (1963, p. 222). H e too subscribes to the notion of play bringing

about the balance required by the ego to function effectively. Erikson describes the play

process as allowing the ego to detach from the situation that caused it fear or anxiety

consequently enabling the ego to maintain its role of mastering and integrating

experience.

A second component of the play process Erikson believed was firmly linked to

the formation of identity and positive self-esteem. H e spoke of the time lag, in modern

industrialised Western 'civilised' cultures, between what he described as " child training

and social actuality" (1963, p. 237), alerting the reader that in such a society there are

few opportunities for children to legitimately experience real accomplishment in a

109


culturally meaningful way. (Bearing in mind those statements were made more than

thirty years ago, it is attractive to imagine Erikson's response to society's present

condition!) Play, Erikson believes, permits children to experience these

accomplishments and successfully nurture a healthy self-esteem and identity.

Importantly, symbolic play emerges when children are simultaneously resolving, or at

least very close to resolving the first of Erikson's proposed psychosocial stages - that of

trust versus mistrust - and experiencing the conflicts inherent in the second

psychosocial crisis of 'autonomy versus shame and doubt' (1963). The medium of

symbolic play permits toddlers to experiment with a range of skills and strategies,

complementing both their growing autonomy and desire for self-preference, in an

almost risk-free environment.

3.7 SYMBOLIC PLAY AS A FUNCTION OF ETHOLOGICAL THEORY

The work of Bruner (1972, 1973) and more recently that of Bjorklund (1997) is

also seen to be pertinent to this study. Bruner's reflections on play are closely aligned to

those theorists w h o adopt an ethological approach to child development. H e emphasises

the nature of the play environment as being a low-stress context where the outcomes of

activity are not needed for survival (1973). Bruner suggested that the value attributed to

play gains support from the Yerkes-Dodson law, which states, "the more complex the

skill to be learned, the lower the optimal motivation level required for fastest learning"

(Bruner, 1974, p. 127). Play is validated in Bruner's description of the play state acting

as a 'buffering environment' in which children could develop skills through trial and

error or combinatorial behaviours upon which the child's survival did not depend.

In keeping with Vygotsky's beliefs, Bruner (1972) suggests that cultural rules

and conventions are transmitted through play. H e also postulates that 'tutor proneness' is

a feature of the young, and supports the notion of 'scaffolding'. This proposal, implying

perhaps the young's apparent readiness to learn in a supportive social context with a

more-experienced partner or facilitator, parallels the Vygotskian view of the importance

of the more-experienced other in aiding the learning of younger children. Importantly,

for the parts of this study (Conditions T w o and Three - mixed-age play and dyad play)

where the impact of the 'more-experienced other' is relevant to the younger child's skill

acquisition, Bruner (1975) considers that the child's being paired with a more-practised

110


other permits the 'joint action' and 'joint reference' crucial to the development of

language or social-skill acquisition.

Consideration of an ethological perspective also gains credence from the work

undertaken by Konner (1975). H e suggests that, historically, in pre-industrial society,

children spent time in play with others of mixed ages, which was likely to favourably

support the younger children's development. H e contends that the types of play -

namely solitary and parallel play - indulged in by infants and toddlers when in a

homogeneous group are to be expected because, until quite recent times, human groups

were too small for children to experience play with same-aged others. Therefore there

has been no evolutionary process in selection of abilities for very young children to

relate successfully to same-age peers (Konner, 1975).

Continuing in this vein, w e note that ethological links arise out of the research

by Bloch (1989) who, although focusing on ecological frameworks of play, point out

that patterns of play both in terms of choice of play partner and in the type of play

exhibited, vary greatly in cultures other than those in 'industrialised countries' of the

largely Western world. Their study, gauging the type of social play engaged in by

American and African children aged two to six years old, indicated that in an African

community children were exposed to a variety of different-aged children most of the

time and consequently rarely engaged in solitary or parallel play. Bloch suggested that

a factor impinging upon the American outcomes of the study could be the limited social

network available to those children, in that much of their time is spent in homogeneous

cohorts (1989). Such findings tend to support Konner's (1975) statements made from an

ethological perspective.

A second link in the Bloch (1989) ecologically based study exists with another

theoretical perspective explored in preparation for the current study. Bloch's findings

indicate evidence of the Vygotskian construct of the more-skilled collaborative other

supporting children's development of a new skill or area of expertise. Although, for the

most part, 94 percent, of social play exhibited by the African children was in the

company of other children, worthy of note was the wide range of adults who were at

various times involved with children in play - albeit briefly.

W h e n contemplating the basic premise of ethology - namely the bio-

evolutionary bases of behaviour - it is tempting to extrapolate parallels in behaviour

between human infants and carers when considering the interesting observations of non-

human primate play made by Biben and Suomi (1993). In keeping with human babies,

111


the first playmate for a young ape (a primate) is his mother. This period is swiftly

followed by the young ape engaging in a phase of solitary play before indulging in

social play with other young members of his species. Interestingly though, infant

gorillas - also a higher primate - prefer older, juvenile play partners to same-age

partners (Fossey, 1979, p. 175), while it seems other species overwhelming choose

peers of a similar age and size as playmates. Observed too was the modification of play

behaviour by the older primates to cater for the younger one's abilities and limitations,

together with the employment of what the researchers term, 'mock behaviour'. This

term describes how the adult primate 'allows' the younger one to 'win' or gain control of

the situation by withholding its true strength or more sophisticated strategies (Biben &

Suomi, 1993, p. 194) - a not dissimilar behaviour to that displayed by human parents in

playing with their offspring.

3.8 SYMBOLIC PLAY AS A FUNCTION OF ECOLOGICAL THEORY

Reference to the work of Bronfenbrenner (1979), an ecological theorist, is also

seen as relevant to this project. The L D C C C s in which children are studied for this

project resemble a component of the cultural ecological system, which Bronfenbrenner

defined as the 'microsystem'. Within this microsystem children are exposed to varied

physical opportunities such as play materials and proximity to children of other ages

and other sexes. The L D C C setting also provides a social network of other people

w h o m children encounter and subsequently interact with, build relationships with and

have in c o m m o n shared experiences.

Bronfenbrenner would consider that the parents' choice of centre, and the degree

to which they are prone, through 'out-of-centre' activities, to extending relationships

and connections made by their children within the child-care setting, to be of

importance. Such activity extends the range of social opportunities available to the

child and would be considered by Bronfenbrenner as being encompassed by what he

terms the 'mesosystem' of his ecological model (1979). Within this context, further

connections between child-care, home and others' homes and families, broaden the

child's social network and opportunities for social activity.

In the early '90s, Bronfenbrenner (1993) reflected upon what he viewed as the

instability in children's lives - partially due to the lack of predictability and consistency

in their daily lives. H e contends that 'the enduring reciprocal interactions' experienced

112


by the child in his immediate environment - for example, between parent and child,

carer and child, and between children themselves - comprise "proximal processes" and,

as such, are crucial to the 'successful' development of the child (1993, p. 8).

Bronfenbrenner emphasises that these reciprocal interactions must take place regularly

over an extended period of time to be effective. In the contemporary situation, children

attending a L D C C C are exposed to a number of adults - one or two of w h o m are likely

to be allocated as their primary caregivers - and a group of usually similarly-aged

children. Such an environment, where children are grouped homogeneously, limits each

child's opportunities to exchange reciprocal interactions within a group of either more-

experienced older peers or less experienced younger peers. If, in L D C C , children are

enculturated into a broader network of children of all ages, there are more opportunities

for all the participants to establish relationships across all age groups to satisfy the

variety of requirements arising out of children's differing needs. Sound research

evidence in the form of enhanced development - for example, toddlers seen to engage

in more developed symbolic play as an outcome of mixed-aged grouping - is likely to

bring about favourable changes with respect to the grouping of children.

In terms of the influences on the social-symbolic play of toddlers in L D C C C s ,

Howes and Unger (1989) who, on this occasion, approached research from an

ecological perspective, suggest that children's play takes place in what they define as

'ecocultural contexts'. They use this term to embrace a variety of human and physical

features in the child-care environment. Considered as being included in the ecoculture

are: caregiver training; play equipment and toys; the adult's role in the easing of

interactions between children; and the provision and organisation of time, space and

materials. Familiarity with peers was also recognised as an ecocultural feature of the

setting. W h e n one considers the current study from an ecological perspective, the

notion of ecocultural aspects of an environment is useful for both identifying and

focusing upon specific attributes of the microsystem.

3.9 CONCLUSION

In this chapter the relevance to the present study has been illustrated of each of

the five theoretical perspectives discussed - cognitive, psychoanalytical, social-

cognitive learning, ethological and ecological. Certainly, the greater part of the research

project is grounded in the work and ideas of the two cognitive theorists, Piaget and

113


Vygotsky. However, it is recognised that it is imperative to be mindful of the insights

offered by the neo-Piagetians.

It would seem that psychoanalytic theory has a clear function in the discussion

of young children's symbolic play and its early use and emergence. There are also

convincing arguments for using this opportunity to take both an ethological and

ecological stance in the furthering of an understanding of the role and nature of toddlers'

emerging symbolic play. As in any thorough and detailed exploration of a subject it is

especially important to note and remark the strengths and limitations of the various

theories investigated.

In the next chapter, Chapter Four, w e turn to an exploration of and discussion

about the concept of play.

114

Chapter 4: The Concept of Play

CHAPTER 4: THE CONCEPT OF PLAY 4.1 INTRODUCTION

This chapter contains a brief outline of the history of play and discusses and

attempts to define play before considering its influence on young children's cognitive

development. As mentioned in earlier chapters, symbolic play (the focus of the current

study) according to Westby (1991) comprises four dimensions or behaviours - each of

which is described in some detail. Brief summaries of the social and cognitive

categories of play, also relevant to the current project, are provided. T w o additional

measures of play are used in this project: the first, the 'level of cognitive effort' required

for play, and the second, the 'thematic unit of play'. Each is explained in this chapter.

(See Appendix Ai for Westby's [1991] Scale for Assessing Children's Pretend Play.)

4.2 THE HISTORICAL CONTEXT OF PLAY

Historically it is interesting to note that playful activity has been both revered

and denounced. The Ancient Greeks viewed play as synonymous with leisure and

seemed of the opinion that such activity was the mainstay of life. Conversely, the

Puritanical movement from the 1500s viewed with disdain any occupation in activity

not seen as related to work (Thut & Adams, 1964).

It is possible that the importance of play in childhood remained unidentified in

Western society until approximately the nineteenth and early twentieth centuries when

philosophers turned their attention to the topic. This statement is supported by

historians such as Aries (1962) who noted that artworks and documents from this period

revealed that society had no concept of childhood as currently perceived and children

were akin to miniature adults and treated accordingly. However, according to

Borstelmann (1983) this assumption may not have been accurate.

Borstelmann postulates that "despite the paucity of domestic behavioural

records" (1983, p. 2) children were indeed recognised as possessing needs different

from those of adults and that childhood was recognised as a specific part of the life

span. Additional evidence m ay be found in a painting 'Children's Games' by the

Flemish artist Brueghel in 1560 which depicts children in the sixteenth century

participating in no less than sixty different children's games, providing evidence

perhaps that some acknowledgement of the importance of play in children's lives was in

existence during that period (Marijnissen & Seidel, 1984).

115


Changes with regard the understanding of childhood were prompted first by the

social philosophers of the seventeenth and eighteenth centuries. Simultaneously,

Froebel (1782-1852) who developed an activity-based program for children (although

intended to be used in a prescribed, sequential manner) highlighted the usefulness and

appropriateness of play as an important component in young children's development

(Connell, 1980, p. 4). A n interest in children learning and acquiring knowledge through

play activity persisted and in the early 1900s Montessori (1870-1952) developed an

educational method "based on freedom for the child in a prepared environment"

(Connell, 1980, p. 134).

Finally, in the 1930s with the emergence of the progressive movement a child-

centred approach to early-childhood education arose (Dewey, 1916; McDonald, 1964).

This movement advocated exploratory, child-directed activity. Opportunities to engage

in free play and the fostering of creativity were promoted by the progressive movement

conducive to young children's development and learning. The constructs perceived by

the progressivists were later expanded by modern theorists including Piaget and more

latterly Vygotsky each of w h o m has added to current knowledge and understanding of

play.

In contemporary times, perhaps the culmination in terms of advocacy for

children's play was seen in the convention of the United Nations' Declaration of the

Rights of the Child (Almy, 1984). This convention considered the child's right to play

equal to the child's right to protection, adequate nutrition, housing, health care and

education.

4.3 DEFINITIONS OF PLAY

The attempt to describe play has generated much discussion amongst those who

tried to define it (Hutt, 1971; Garvey, 1990; Bruner, 1977). The result was that to some

degree a consensus was reached that play is a pleasurable, spontaneous behaviour,

child-directed and therefore intrinsically motivated and usually involves "active

engagement" (Garvey, 1990, p. 10). Play often has an exploratory nature (Hutt, 1971)

and children m ay be engaged in play alone or with others. Children may, in play,

incorporate specific play materials or other materials not initially recognised or intended

as play materials. Interestingly, Bruner (1977, p. 5) differs from other theorists in that he

considers play, "an approach to action, not a form of activity". The implication in this

116


statement is that children do not indulge in play as a means to an end or as a process

resulting in outcomes, but engage in play for the inherent pleasure. Bruner (1977)

postulates further that play is a means for both exploration and invention. Elkind

(1981b) has suggested that play is the antidote for what he perceives as the inordinate

amount of stress experienced by many young children.

4.4 THE INFLUENCE OF PLAY ON DEVELOPMENT

M a n y theorists and researchers (Parten, 1932; Piaget, 1962; Vygotsky, 1978;

Johnson & Yawkey, 1988; Johnson, 1990; Smilansky & Shetfaya, 1990) have attempted

to identify the influence of play on the development of young children. While attention

has been focused on the likelihood of play positively affecting change in each of the

developmental domains, the areas of social and cognitive development remain those

most often claimed as benefiting from play experiences. A surfeit of such claims has

been reported in the last forty years.

Conclusions drawn by Fisher (1992), who conducted a meta-analysis of 46

studies involving play, were useful for both setting this discussion of play into context

and establishing the parameters for the range of literature reviewed in Chapter Four.

Fisher (1992, p. 172) concluded that a category of play he identified, as "child-

orientated" play is the single most effective type of play with the potential to positively

influence children's cognitive, language and affective-social development. H e described

such play as imaginative play - a type of play requiring children to employ the skills of

object and situation transformation as is required for engaging in symbolic play - the

focus of this study.

4.5 SYMBOLIC (OR PRETEND) PLAY

To successfully engage in symbolic play, the children are required to represent

objects, persons, actions and places symbolically. Garvey describes this as occurring

when the child "acquires the ability to encode his experiences in symbols;" (1977, p.

14). Simple symbolic play, which young children develop in the second year of life, is

thought to be the precursor to a more complex type of symbolic play - thematic pretend

play - currently of interest to play researchers. Thematic play, which is a feature of

preschool-aged children's play activity, is epitomised by the players exhibiting, through

117


play, activities and knowledge that are unrelated to their lives. Thematic pretend play

would seem to have a positive role in several areas of cognitive and social development.

It appears, for example, that a reciprocal relationship may exist between thematic play

and divergent problem solving (Wyver & Spence, 1999). If, as it seems appropriate to

suggest, symbolic play is a forerunner to thematic fantasy play,, then delays or

impairment in the onset and subsequent development of symbolic play may hamper the

emergence of and children's progression into thematic fantasy play and the ensuing

cognitive gains such as divergent problem-solving skills.

As mentioned earlier in Chapter One, several researchers uphold the notion that

symbolic play in itself is supportive in the development and promotion of cognitive

skills (Rubin, 1980; Vandenberg, 1981; Westby, 1991). Smilansky (1968), identified

symbolic play as a mechanism through which young children gained an understanding

and knowledge of others (and their perspectives) in their environment. This ability,

Mead (1932) considered then aided children's own development of 'self-knowledge and

understanding.'

4.5.1 Characteristics of Symbolic-Play Development

The emergence and subsequent development of symbolic play is itself a

developmental process. Four behaviours or, as described by Westby (1991),

'dimensions', comprise symbolic play - 'decontextualisation', 'thematic content',

'organisation of themes' and 'self-other relationships'. Children's behaviour in each

dimension changes and increases in complexity as they both mature and develop

cognitively.

Each of the four symbolic-play behaviours or dimensions is described below.

4.5.1.1 Dimension One: 'Decontextualisation'

The type of props used by children in symbolic-play activity identifies the level

of 'decontextualisation' evident in their play. These props m a y be classed as realistic

and life-like as is likely in earlier levels of symbolic play. At more complex levels,

players are likely to use either replicas of real objects or another object instead of the

desired prop - incorporating a representational component to their pretence. It is likely

too that at these higher, more complex levels of symbolic play children will use

materials, for example building blocks, to construct three-dimensional structures to

support their play and may also incorporate verbal descriptions to further 'set the scene'

118


for their pretend episodes (Westby, 1991). The earlier level of 'decontextualisation'

where life-like props are required for use, is applicable for the first five levels of

symbolic play - the change in prop usage likely (according to Westby, 1991), to appear

when children are approximately three to three and a half years old or more (see

Appendix Ai). As children's cognition increases, so their dependence on props with

life-like qualities lessens and the use of more abstract and finally imagined props comes

into being.

4.5.1.2 Dimension Two: 'Thematic Content'

In keeping with developmental changes, modifications in children's use of

thematic content are also observable. Initially in young children's symbolic play, their

schemas and scripts revolve around familiar, everyday events and activities in which

they have been involved. Activities of familiar others are then enacted as symbolic

activity increases in complexity, and toward the end of the toddler period, at

approximately 30 months of age, children are able to represent events which have not

been regularly experienced and so have a rather novel quality to them (Westby, 1991).

Expertise in this dimension of symbolic play ('thematic content') may support the

development of thematic pretend play. This category of play (referred to earlier in

Chapter One) features imaginary and fictitious themes and is thought to be an important

component of preschoolers' play activity since it correlates with a range of cognitive

skills.

4.5.1.3 Dimension Three: 'Organisation of Themes'

Young children's organisation of scripts and schemas undergoes similar changes

as both development and maturity impacts upon their symbolic play. The very

beginnings of symbolic play, often seen emerging from functional or practice play,

involves toddlers engaging in single, brief and isolated pretend actions and although

multiple actions may be represented it is likely that these will be quite unrelated to one

another. In time, two actions are then combined leading to what Westby termed

"elaborated single schemes" which toward the end of the toddler period increase in their

complexity to become "evolving episode experiences" (1991, p. 137).

119


4.5.1.4 Dimension Four: 'Self-Other Relationships'

In the initial stages of symbolic play children's pretend actions are conducted

upon the children themselves, in that they are the agent for the pretend actions or

behaviour. Westby terms this practice of single behaviours such as pretending to drink

from a cup or eat from a fork as "autosymbolic or self-representational pretend play"

(1991, p. 141). At this stage decentration has not occurred and no others are involved in

the symbolic play. As their symbolic activity develops, toddlers' actions incorporate

either objects or people, since the behaviours indicating the presence of symbolic play

now contain an 'acting upon' component. Such behaviour reflects the process of

decentration occurring and the advent of 'self-other relationships', each of which signals

children's emerging knowledge and awareness of first, others' perspectives and,

secondly, the interrelationships that exist between objects and people. These

developments in young children's thinking and knowing constitute a marked qualitative

change in children's cognitive processes.

4.6 SOCIAL PLAY

Though not a focus in the current study it is thought useful to note the type of

social play engaged in by participants. (This will also provide additional social-

contextual information.)

The nature of simple social symbolic (or pretend) play - the focus of the present

project - involves two children interacting in play. Parten's (1932) research aiming to

identify the complexity of social play among participants and their peers, involved

children aged from two years to approximately five years old. Her findings resulted in

the development of a scheme describing six categories for rating social play.

Recent research by Howes (1991) found that children w h o entered L D C C early,

as infants and toddlers, were more socially competent with their peers than children who

entered child-care later, as preschoolers. These 'early-entry' children also spent more

time playing than 'late-entry' children (Howes, 1991). While maternal attachment was

also a component of the study, Howes suggested that the social context of L D C C ,

especially where the quality of care-giving is high, might be more influential on young

children's acquisition of social skills than the mother-child relationship.

Parten's (1932) categories of social play categories are described in more detail

below.

120


4.6.1 Categories of Social Play

4.6.1.1 Unoccupied Play

Activity undertaken by children involving either standing around or moving,

from one area to another, in what appears to be an unplanned, random manner,

seemingly without a definite purpose, is the first category of social play identified by

Parten (1932) - that of unoccupied play.

4.6.1.2 Onlooker Play

The second of Parten's (1932) categories, onlooker play, occurs if the child's

gaze and physical presence is directly toward others engaged in some activity. Such

behaviours m a y be supported by other non-verbal behaviours indicating an onlooker

role. Onlooker play is relatively easy to identify as children's proximity assists the

researcher's decision. For activity to be rated as onlooker play, children are usually

well within a metre of the children they are observing playing or occupied in some

activity. Often children engaged in onlooker play move closer to the observed group or

individual and display a discernible degree of intensity of gaze toward the group or

other player.

4.6.1.3 Solitary Independent Play

This type of play, characterised by children playing alone even if others are

within close proximity, is a very c o m m o n feature of toddler play. Solitary play often

appears to fulfil specific requirements - for example allowing opportunities for practice

play to occur, as is often the case in physical play. It also offers opportunities for

children to withdraw from the group environment and enjoy quiet, private play periods.

In group situations, such as those encountered by young children in L D C C C s , very

young children become skilled at engaging in solitary play while to the casual observer

they appear to be part of the wider larger group.

4.6.1.4 Parallel Play

Of the child engaged in parallel play, Parten commented, "he plays beside rather

than with other children" (1932, p. 250). This category includes children who share

materials and space and w h o are often in close proximity with one another without

influencing the activities of others and without interaction occurring. Identification of

parallel play requires close observation, as solitary play can and often does occur within

121


the same setting. It is possible for several children to be engaged in parallel play, but in

many instances a child can be quite apart from the nearby group, although sharing

materials and space, and can indeed be solitarily occupied. Again markers such as gaze

and physical placing of body aid the observer in categorising this type of social play.

4.6.1.5 Associative Play

Parten (1932) described activity displayed by a group of children sharing

materials and space and occupied in similar play, while no real coordination towards a

common goal is apparent, as associative play. Children's behaviour, when engaged in

this type of social play, m a y indicate a degree of cooperation in that components maybe

exchanged or passed over, but projects are likely to be individual and not subscribed to

jointly. Associative play is rarely evident in toddler play, but can be initiated by older

peers w h o have the language skills necessary for implementation and subsequent

maintenance of such play.

4.6.1.6 Cooperative Play

The most sophisticated and complex of the range of social categories of play

described by Parten (1932) is cooperative play. Players occupied in cooperative play

manifest behaviours such as cooperation, negotiation and joint construction.

Imaginative, pretend and fantasy themes are evident together with role taking, role

allocation and turn-taking in leadership, consistent with the player's ability to sustain

play over a period of time or even days.

4.7 COGNITIVE PLAY

Cognitive play, like social play, is not a focus of the present project but is an

important consideration in any discussion of the concept of play. Seven types of

cognitive-play - non-play, sensori-motor, exploration, physical play, functional or

practice play and constructive play - are described briefly below. (As was done with

social play the type of cognitive play engaged in by participants will be noted.)

122


4.7.1 Categories of Cognitive Play

4.7.1.1 Non-Play

This category describes activities undertaken by the children that are either

teacher-directed or teacher-assigned - for example a group story or washing hands

before morning tea. These and similar activities are designated 'non-play'. Activity

exhibited by children during unoccupied play, perhaps walking, wandering in or moving

through the play environment, or maybe transiting from one activity to another, would

also be classed as non-play. In summary non-play indicates an "absence of behaviour

recognisable as play or exploration" (Wyver & Spence, 1995, p. 43).

4.7.1.2 Exploratory Play

Although it appears there remains a lack of consensus among researchers with

regard to distinguishing between play and exploration, the definition of exploration for

the purpose of this project is in keeping with the findings of Garvey (1977) and Rubin,

Fein, and Vandenberg (1983). These researchers' notion that the difference between

exploration and play is the "locus of control" (1983, p. 732) - namely whether it is

external or internal - is aptly illustrated by Garvey's description of the movement from

exploration to play as "A new experience . . . then exploration . . . treated more lightly

and enjoyed" (1990, pp. 32-33).

Exploration of an object is seen as an information-gathering exercise that may

precede subsequent playful activity involving that object. The very nature of

exploration decrees that it occurs during solitary play - the third category of social play

described by Parten (1932). Children, having explored an object, may then use it in

functional or practice play as the transition, from "what does this object do" to "what

can I do with this object" occurs (Hutt, 1971, p. 246). Children engaged in exploratory

play often employ perceptual motor activity in a manner described by Wyver and

Spence (1995, p. 44.) as "stereotyped perceptual motor examination".

4.7.1.3 Physical Play

A n example of physical play would be the climbing on apparatus arranged

specifically to encourage the development of gross-motor or large-muscle skills.

However, activities such as trike-riding or digging in the sandpit would be classified as

practice or functional play.

123


4.7.1.4 Practice or Functional Play

Repetitive activity, with or without objects was termed functional or practice

play by Piaget (1962). Importantly this type of play is characteristic of children's

activity as they progress toward the conclusion of what Piaget described as their

sensori-motor stage of cognitive development occurring in the first two years of life.

Such play allows children to engage in a variety of behaviours, the development of

which m a y or may not be promoted and supported by the available equipment and

materials, for example sand play with implements or a repeated movement sequence.

The pleasure experienced by a child involved in functional play activity is reflected in

Piaget's words "for the mere joy of mastering it and showing off to himself his own

power of subduing reality"(1962, p. 162).

One signifier for symbolic play occurring during functional or practice play is

'decontextualisation' - Dimension One in Westby's (1991) scale - whereby a life-like

prop is used to represent actions or events children have themselves experienced. (This

dimension of symbolic play was explored more thoroughly earlier in this chapter, 4.7.1.)

4.7.1.5 Constructive Play

As the name suggests, constructive play produces a product or end result, in that

something concrete exists after children's play with components or materials is

completed. Children in constructive play strive to make a representation of an object or

situation. Van Hoorn, Nourot, Scales, and Alward suggest that constructive play

provides "a natural link between functional play and more sophisticated forms of

symbolic play" (1993, p. 35).

4.8 COGNITIVE EFFORT

A number of researchers have given consideration to the amount of cognitive

effort or input required to execute a play activity. Aureli and Colecchia (1996) explored

the "Test-Operate-Test-Exit" (TOTE) cycle of operation suggested by Miller,

Galnanter, and Pribram (1960, p. 27) as being part of any activity episode including

episodes devoted to playful activity. The T O T E process requires the player (or agent)

to plan toward a known goal, employ various actions to achieve this goal and, once the

goal was achieved, to stop the activity. The components of the T O T E process perhaps

support or even parallel the notion of intentionality - an idea further extended by Aureli

124


and Colecchia (1996) to form the basis of an instrument to identify the cognitive level

employed by children to carry out an episode of play activity.

Intention, utilisation and completion were the three criteria that Aureli and

Colecchia (1996) defined for identifying the level of cognitive requirement in a play

episode. The child collecting or preparing the necessary materials or means toward

executing the play activity indicated intention. Utilisation, the second criterion, was

recognised when the child made appropriate advances towards accomplishing the

activity by combining or arranging the materials or objects in some manner indicating

progress towards the end goal. Finally, completion was noted when the child indicated

(usually verbally) that he or she had achieved his or her planned goal. In rating the level

of play Aureli and Colecchia (1996) stipulated that for a high level of cognitive

requirement all three criteria were to be present. Children's play activity reflecting one

criterion was rated as medium level in terms of cognitive input while play activity in

which none of these criteria was present was described as low level.

In this study it was deemed as appropriate to code the observations for level of

play in terms of cognitive requirement, and so the strategy of identifying whether the

play episodes were of high- medium- or low-level cognitive input was employed in

accordance with Aureli and Colecchia's use of the rating scheme (1996).

4.9 THEMATIC UNITS OF PLAY

A second instrument, used by Aureli and Colecchia (1996) in the previously

mentioned research, measured the length of time children engaged in a thematic-play

episode. These researchers suggested that any play activity lasting more than one

minute could be regarded as thematic. Therefore children's performance in one activity

on the same theme for two or three minutes was to be rated as a brief thematic unit,

while involvement of at least four minutes was to be classified as a long thematic unit.

4.10 CONCLUSION

In this chapter a range of concepts relating to play were explored. These are

central to the understanding of play as a concept underpinning this project. In summary

it can be seen that play is largely child-directed and clearly developmental in nature.

Symbolic play - the focus of this study - is thought to support and promote children's

125


understanding and cognitive development (Nowak-Fabrykowski, 1994; Singer, 1994;

Wyver & Spence, 1995; Chaille & Silvern, 1996; Dockett, 1998). Certainly the

emerging symbolic play toddler participants are expected to exhibit is a precursor to the

more complex forms of symbolic and thematic play. Results of this study might help

indicate whether toddlers' free play with older peers (in either group mixed-age free

play or dyad free play) in the L D C C environment favourably influences early symbolic-

play development.

The following chapter, Chapter Five, presents the research method employed for

the current enquiry.

126

Chapter 5: Methodology

CHAPTER 5: METHODOLOGY 5.1 INTRODUCTION: THE PURPOSE OF THE PRESENT RESEARCH

It has been explained that the purpose of the current research is to identify the

influences, during mixed-age free-play periods and dyad free play, of older preschool-

aged peers on toddlers' symbolic play. Interest in the impact or influence of older peers

when interacting with younger peers is reflected in a considerable amount of research

conducted over the past two decades (Goldman, 1981; Howes & Farver, 1987; Blasco,

Bailey, & Burchinal, 1993). Symbolic or pretend play requires the player to represent

events, objects or situations when apart from the reality of each - described so aptly by

Piaget as ". . . applications of schema to inadequate objects and evocation for pleasure. .

. " (1962, p. 97)

The present study is concerned with cognitive outcomes in terms of the

frequency and level of toddlers' symbolic play.

5.2 HYPOTHESIS AND ASSOCIATED QUESTIONS

As stated in Chapter One, (1.4), the major hypothesis is, that toddlers engaged in

free play with older peers (in either mixed-age play or dyad play) will exhibit symbolic-

play skills in each of the four dimensions of symbolic play identified by Westby (1991)

- 'decontexutalisation', 'thematic content', 'organisation of themes' and 'self-other

relationships'- more frequently and at a higher level, than when they are engaged in


In conjunction with the major hypothesis, the following questions will also be

considered:



symbolic play?


of the four dimensions of symbolic play identified by Westby (1991) -

'decontexutalisation', 'thematic content', 'organisation of themes' and

'self-other relationships' - more frequently and at a higher level than


c) Does the L D C C environment, which includes the quality of care

provided - the program of activities implemented, the materials and

127


equipment provided, staff training and qualifications, and other available

resources - support children's engagement in symbolic play?

5.3 RESEARCH METHOD

5.3.1 Influences on the Research Design

As described in the literature reviewed in Chapter Two, a number of researchers

(Howes & Farver, 1987; Mounts & Roopnarine, 1987; Roopnarine et al., 1992; Blasco

et al., 1993; Bailey et al., 1993; Lyytinen, 1995; Dunn, Kontos, & Potter, 1996) have

found a range of different outcomes resulting from very young children's engagement in

mixed-age play with older peers. (Most of these investigations took place in LDCCCs.)

More powerful findings were shown when the younger child was paired with an older

peer in a dyad situation (Howes & Farver, 1987; Lyytinen, 1995). (In the studies cited,

allocation of the older peer was random.) Where symbolic (or pretend play) was the

dependent variable findings indicated that in mixed-age play the older peers had the

skills to structure the roles for themselves and the younger peers who, exhibited more

complex cognitive play (Howes & Farver, 1987). Dunn et al. suggest that interactions

resulting from younger children's engagement with older peers are "powerful predictors

of development" (1996, p. 349). These findings cited above inform the major

hypothesis for the present enquiry that incorporates the premise that older peers can

have a significant role in supporting and scaffolding the younger children's

development of more complex symbolic play. Additionally, a component of the study

conducted by Dunn et al. which also has a bearing on the current research, was their

observation that older peers featured more as play partners for the younger children

when the care-giving environment was of a lower quality and contained less responsive

caregivers.

The dyad condition whereby a younger, toddler-aged child is paired with a

familiar and where possible 'self-chosen' older peer would seem to be a particularly

useful social context for the younger peer's learning and skill acquisition. Such a

situation has the potential to fulfil the requirements for a number of theoretical

propositions including: Vygotsky's concept of the 'zone of proximal development';

Rogoff's (1990) notion of 'guided participation'; Tomasello's (1995) concept of 'joint

attention' and Schaffer's (1996) idea of 'joint involvement episodes'. Central to each of

128


these concepts is the notion of two participants in a shared context with a common

orientation or point of reference.

Importantly studies investigating the influence of older peers upon younger

children are often conducted by way of naturalistic observations of children in their

familiar settings (Howes & Farver, 1987; Blasco et al., 1993; Dunn et al, 1996).

Familiar settings in which research has been undertaken include children's own family

homes, infant rooms in L D C C C s , Family D ay Care homes, preschools and primary-

grade classrooms in schools.

5.3.2 The Research Method for the Present Project

The current research project is best described as a semi-naturalistic study - the

method employed supporting both ecological and ethological validity. Shaffer (1985)

cautions the naturalistic observer to be wary of subjectivity and observer bias. H e

suggests that an effective method of counteracting these two factors is first to "specify

in advance precisely what kinds of activity quality as examples of behaviour you wish

to study" (1985, p. 16), and secondly to incorporate a reliability process. Westby's

(1991) Scale for Assessing Children's Pretend Play was used in this present study - this

fulfils Shaffer's first caution and a stringent inter-rater reliability process involving two

academics (each skilled in the area of play content) fulfils the second.

Additionally the current researcher recognises that in a naturalistic environment,

both natural events and the participants within that environment are largely

'uncontrolled'; therefore study findings m a y be less precise (indicating direction as

oppose to cause and effect) than those resulting from a more controlled experimental

situation (Shaffer, 1985).

In order to collect semi-naturalistic data for the current project, toddlers

attending L D C C C s were observed and video-taped while engaged in play in three

conditions. In Condition One, participating toddlers were observed in free play with

their same-age peers in an outdoor play environment. Condition T w o required the

toddlers to be observed in free play with their older preschool-age peers in the outdoor

play environment - exactly the same environment as that shared with their same-age

peers in Condition One. The third condition was dyad play where an older peer was

required to engage the toddler in play. The decision to include a dyad play condition

(Condition Three) arose from the literature reviewed in Chapter T w o (2.2) showing that

129


older peers favourably influence younger children's play - findings that have been

especially strong in dyadic conditions.

Condition Three (dyad play) was perceived as a 'vignette' or an 'insightful

view' of what possibilities may exist (but may not have eventuated) for the younger

child in the mixed-age play condition. The premise underlying dyad play was that the

older peer would be likely to engage in interaction with the toddler as a more-

experienced player. The prediction was that the situation presented an opportunity for

the younger child to move within his or her 'zone of proximal development' aided by

the more-skilled older peer. In addition it was expected that the older peer would

engage in communicative behaviours supportive of the toddlers' activity and that the

younger child might engage in imitative behaviours.

The premise that the older peers would use communicative behaviours likely to

sustain the toddlers' pretend play seemed to be supported during observations made

before the commencement of the research project and during the period in which the

methodology and instruments were piloted. For example in the dyad condition when

play dough was the play medium the older peer suggested to the toddler partner, "You

go to school and I'll make a cake. You can eat it when you come home." As the

toddler hesitated the older peer said, "Just go round the table and come back home."

The toddler then complied with the older peer's request (Researcher's field notes).

In summary, the toddlers had two opportunities to engage in play with their

older preschool-aged peers: in Condition Two, a mixed-age play session, and in

Condition Three, a dyad-play situation. The materials available to the children in the

first two conditions were those supplied by the L D C C C staff in accordance with the

planned centre program, while the researcher supplied the play materials for Condition

Three. The three conditions are discussed in more detail later in this chapter in 5.6.2.6

and the provision of materials in 5.6.2.7.

The present study is concerned with cognitive outcomes in terms of the

frequency and complexity of toddlers' symbolic play. As foreshadowed earlier in this

chapter (5.2) it is postulated that toddlers will, in their play with older preschool-aged

peers (in two conditions - mixed-age play and dyad play), first exhibit symbolic-play

acts (or events) more frequently and secondly display symbolic-play skills at a higher

level than expected for their age - according to Westby's (1991) scale.

130


5.3.3 Ethics Approval

Ethics approval was applied for and subsequently granted by the University of

Wollongong, first, for the implementation of the present study and secondly, to trial the

observation procedures late in 1994. (The project itself began in early 1995.)

Institutions administering the L D C C C s were each approached and the University of

Wollongong ethics application together with all documentation to be used with parents

and staff were submitted for approval to the boards or committees of the various

governing bodies. T w o of these also required their own Ethics Committee to approve

the data-collection procedure. In each case the procedure was approved. In summary

all applications received approval and no modifications to the documentation were

requested.

5.3.4 Verification of Methodology and Instruments

While not constituting a pilot study it was thought wise to trial methodological

components and procedures to first, ensure that they operated smoothly and in keeping

with the expectations and secondly, so any modifications indicated could be

incorporated before the study began.

Four L D C C C s known to the current researcher were approached and the

procedures as outlined in this chapter in 5.6 and 5.7 were followed. These included: all

documentation to staff and parents; observation and video-taping of the three free-play

conditions (same-age play, mixed-age play and dyad play); provision of materials for

dyad play; and coding of video-tapes for symbolic play. In addition a university

colleague trained the present researcher in the administration of the Peabody Picture

Vocabulary Test - Revised, Form M (1981).

At the conclusion of the trial period it was decided that minor modifications

were required to enable the distribution and collection of all documentation for staff and

parents to occur with minimal involvement of L D C C C staff. (This was addressed

through the use of collection boxes and 'return to sender stamped addressed

envelopes'.) Additionally minor adaptations were made to the play coding sheets to

enable easier completion. (Examples of all documentation used in the project are found

in the appendices, references to which occur later in this chapter.)

131


5.4 SELECTION OF LDCCCs

Centres in close proximity to the main business areas of Sydney, a city of four

million inhabitants, were identified. Care was taken to ensure an equal distribution of

centres across the north, south, west and eastern inner suburbs of this extensive city was

used in this study. Using a city m a p concentric circles were drawn around the central

business area and L D C C C s in each geographical quarter were approached to take part in

the study. If an approach was rejected the next L D C C C (provided the daily program

met the researcher's requirements) closest to the designated area was approached.

5.4.1 Criteria for LDCCC Selection

Criteria used in selecting the L D C C C s included the proximity to several high-

density business districts (confirmed by the selection process) and the provision of at

least one mixed-age free-play session and one same-age free-play session in the outdoor

environment per day. Familiarity with L D C C C programs allowed this current

researcher to recognise that play outdoors is classified as a 'free-play' period and, for the

greater part, play outdoors reflects both child-initiated and child-directed content. For

those reasons it was important that this free-play session was held outdoors and that the

toddler participants also had the opportunity during the day to engage in same-age free

play in that same environment.

The location of the L D C C C s (all within a nine-kilometre - or approximately

five-mile - radius of the city) ensured, to some degree, that children would be those of

working parents and be in care for approximately eight hours a day for several days a

week. Attending children were also likely to be present at the centre either early in the

morning or late in the afternoon, which is often the time when mixed-age grouping

occurs.

While substantial external validity cannot be claimed in a semi-naturalistic study

such as this one, nevertheless, in order to make the study as representative as possible, a

range of L D C C C s reflecting a cross-section of those organisations providing long-day

child-care was included, hence the inclusion of centres operated by private owners, local

councils, corporate institutions and the community groups. Of the twelve L D C C C s

used in the study, two were operated by local government (municipal councils), three

were privately owned and managed, three were operated by a children's services

institution (two of which were work-based centres in collaboration with privatised

132


utilities, the third being community-based) and four were run by community groups. Of

the last category one centre also provided a work-based service.

W h e n the staff of a L D C C C expressed interest in being involved in the study a

letter (see Appendix Bi) accompanied by examples of all other relevant documentation

was sent to the Director. Staff information letters and staff consent forms (see

Appendices Bii and Biii) were distributed at least three weeks before observations were

scheduled to begin. Parent information letters and consent forms (see Appendices Ci

and Cii) were distributed at the same time.

5.4.2 The Quality Improvement and Accreditation System

Following an interim accreditation measure established in 1991, the National

Childcare Accreditation Council was established in 1993. Its brief was described as

being to oversee the implementation of the Commonwealth government's 'Quality

Improvement and Accreditation System'. This system required the participation of all

L D C C C s that were in receipt of Childcare Assistance.

The process of acquiring accreditation requires the L D C C C to meet standards in

the following areas:

- the interactions between staff, children and parents;

- the centre program;

- nutrition, health and safety practices; and

- centre management and staff development (Wangmann, 1995).

Of the L D C C C s involved in the current project, one centre from each category

of operation (council, private and community) was accredited for the maximum period

of three years. Others ranged from achieving one years' accreditation to failure to reach

a minimum accreditation standard. As can be seen in Table 5.1 one newly opened

centre had not yet been required to participate in the accreditation process.

133


Table 5.1

Characteristics of Participating L D C C C s

L D C C C Id

1

2

3

4

5

6

7

8

9

10

11

12

Category of L D C C C

Children's services (work-based)

Children's services (community)

Municipal

Community

Private

Private

Municipal

Community

Community

Private


Community (work-based)

Accreditation (1, 2 or 3 years)

3

3

3

1

3

process not yet undertaken

3

3

failed accreditation

3

3

2

5.4.3 The Effect of the Removal of Child-Care Subsidies

Shortly after data collection began in 1997, the current Federal government

removed the operational subsidies for community-based L D C C C s and reduced the

child-care cash rebate. These actions, possibly in conjunction with a general economic

downturn, resulted in many parents either withdrawing their toddler-aged children from

L D C C C s or reducing the number of days their children attended. Other effects of

funding withdrawal included children, more often the younger children in the family,

being placed in informal care, or a reduction was made in the parents' hours of work

enabling parents to care for their children. It was reported by the Council of Social

Service in N e w South Wales (NCOSS) that a state-wide telephone 'phone-in' survey

revealed that more than half of the 587 respondents had moved to using informal care

arrangement for their children for at least part of the week (NCOSS, 1997).

Anecdotally L D C C C staff reported that parents often undertook these measures for their

younger, toddler-aged children in preference to tampering with the attendance of an

older, preschool-aged sibling. As could be anticipated, this course of events severely

134


reduced the number of participants fulfilling the criteria for inclusion in the current

research.

5.4.4 Features of the LDCCC Play Environment

The number of children present during the observation periods varied widely. In

part this was due to the number of children enrolled in each centre. However, variations

were also partially dependent upon the time of day and the day of the week the

observations took place. For example some L D C C C s had more children present in the

morning mixed-age play session than the one held in the afternoon. M a n y centres had

one day in the week (often a Monday or a Friday) where numbers of children attending

were substantially lower than on other days. The number of children in the observed

periods ranged from 12 to 30 in a mixed-age group and 5 to 15 in a same-age group

free-play session. Usually each group (mixed-age and same-age) had an outdoor free

play session twice a day - one early in the morning (generally before mid-morning) and

another late in the afternoon. W h e n formulating the research method the goal was to

compile an equal number of video-taped observations in the morning and afternoon in

each L D C C C . This, however, proved to be a logistic difficulty since both toddlers' and

older peers' attendance (and participation) in the late afternoon was sometimes

irregular. For example parents might collect their children earlier than was expected, or

children had perhaps slept for an extended period of time and were involved in routine

matters (for example a nappy change or afternoon tea) and therefore not taking part in

the regular afternoon program. Hence it was necessary for the researcher to make a

number of repeated visits to the L D C C C s to maintain group stability. Nevertheless, 28

percent of the mixed-age play observations were conducted in the afternoon - with 72

percent collected in the morning - and, whilst 77 percent of the same-age play

observations were gathered in the morning, 23 percent were obtained in the afternoon.

Before each filming session the number of staff and children present in the

environment was recorded. Comprehensive field notes detailing the toys, materials and

equipment available for the children's use in the outdoor free-play environment were

also kept. These records enabled the child-care arrangements, including ratios of staff

to children and the provisions for play in the outdoor area, to be documented and

considered for the purposes of augmenting data analysis.

135


5.5 S A M P L E

5.5.1 Recruitment of Toddler Participants

Forty-eight toddler and thirty-seven preschool-aged participants were recruited

from the 12 participating LDCCCs in the inner-city region of Sydney a city with a

population of approximately four million. Participants' ages ranged from 17 months to

63 months.

Information letters were distributed to all staff and parents and the researcher

also attended staff and parent meetings in order to answer any questions about the

project. Informed consent was gained from the parents of all children attending the

LDCCCs as those other than the targeted participants were likely to be incidentally

video-taped in the course of their activities. Centre staff also gave informed consent.

As mentioned above, 48 toddlers were involved in the study. Toddler

participants were required to be aged between approximately 18 to 30 months, to have

both attended the L D C C C and participated in the mixed-age free-play sessions for more

than three months. In addition participants were required to attend the centre's daily

program for a minimum of three days per week. Approximately 80 percent of those

toddlers identified by the researcher as fulfilling the criteria were granted parental

consent.

Although 48 toddlers have featured in this study it was only possible to obtain

complete sets of data (approximately 90 minutes observation under three different

conditions) for 24 of those children. Therefore the sample for this study comprises 24

toddlers. The children, eleven boys and thirteen girls, ranged in age from 17 to 31

months (SD = 4.55). Ninety-six percent of the children spoke English as their first

language. One child, for whom English was a second language, was included as a

participant as the centre staff reported that he was able to communicate effectively in

English to both staff and children at the LDCCC. Characteristics of the toddler

participants are presented in Table 5.2.

136


Table 5.2

Characteristics of Toddler Participants

Child Id

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

Sex

F

M

F

F

M

M

M

F

M

M

F

F

M

F

F

M

F

M

F

F

F

M

F

M

Age in months

17

17

17

18

19

20

20

20

21

21

22

22

24

25

25

25

26

27

28

29

29

30

30

31

Birth order

Second

First

Second

First

Second

First

First

First

Second

First

Second

First

Not known

First

Second

First

First

Third

First

First

First

First

Second

Second

N siblings, sex and age in

months

Male 59

None

Male 67

None

Male 33

None

None

None

Male 81

None

Male 95

None

Not known

None

Female 51

None

None

Female 113 Female 79

None

None

None

None

Female 80

Female 39

N months enrolled

in LDCCC

3+

12

3+

6

6

6

3

12

14

8

13

3+

3+

6.5

5

3

3+

18

18

5

12

6

12

5.5

Ndays LDCCC Attended

3

5

3

5

4

3

3

4

3

5

4

5

5

5

3

5

4

3

5

4

5

5

5

3

137


It was necessary to observe children aged approximately 18 to 30 months

because of changes usual in symbolic play at about 24 to 30 months (McCune-Nicolich,

1977; Westby, 1991.) At the earlier age (18 months) with the onset of symbolic play,

the child is the agent (the 'doer') and using life-like props engages in deferred imitation

involving single acts of pretend play targeted toward him or her self. Later at about 30

months of age the child is able to use substitute props to represent a wider range of

familiar themes - often engaging objects or others as agents in a logical sequence of

symbolic actions. Vocalisation or expressive language is also a feature of the older

toddler's pretend play.

In summary: the mean age of participants at the time of data collection (video

taped observations) was 23.5 months; the mean age of participants' entry into L D C C C

was 15 months; the mean length of time participants had attended L D C C C before data

collection commenced was approximately eight months; and the mean number of days

per week participants attended L D C C C was four. This information and the ranges from

which the means were calculated are displayed in Table 5.3.

Table 5.3

Summary of Toddler-Participant Information

Age at time * Period in L D C C C Weekly L D C C C of observation before data attendance

collection

Mean 23.5 months 8 37months

Range 17 to 31 months 3 to 18 months

Standard 4.55 4.38 deviation

* denotes a very approximate figure because for six participants the only information available for length of attendance in L D C C C was 'more than three months'

5.5.2 Characteristics of Toddler-Participant Groups

The research design contained four experimental groups - Groups A, B,

C and D. The characteristics of each group are as follows:

- Group A Five males aged approximately 17 to 31 months

(SD = 4.77) - each with an older sibling.

- Group B Five females aged approximately 17 to 31 months

4 days

3 to 5 days

8.99

138


(SD = 5.54) - each with an older sibling.

- Group C Six first-born males aged approximately 17 to 31

months (SD = 4.62).

- Group D Eight first-born females aged approximately 17 to 31

months (SD = 4.20).

These details are presented below in Table 5.4.

Table 5.4

Characteristics of Toddler-Participant Groups

With older sibling Without older sibling

Condition Males Females Males Females

Group A Group B Group C Group D

1. Same-age 5 5 6 8

play

2. Mixed-age 5 5 6 8

play

3. Dyad play 5 5 6 8

5.5.3 Recruitment of Older Preschool-Aged Peers

Thirty-seven preschool-aged peers participated in the current study - 11 less

than the 48 toddler participants. This discrepancy in participant numbers is the result of

several factors. First, the dyad-play condition was the final stage of the data collection

and in 11 cases the data collections were discontinued before any of the dyad-play

situations were observed - hence no older peer was identified for involvement with

those toddler participants. Reasons for discontinued observations included: the

withdrawal of toddler participants from the L D C C C ; a reduction in the toddlers' days of

attendance to less than three days per week; and the toddler's absence from L D C C C due

to unexpected lengthy holidays or periods of sickness. In 13 cases, although

observation of dyad play had commenced, similar reasons to those cited above

(pertaining to either the older peer or the toddler) resulted in incomplete sets of data.

139


Consequently while 37 preschool-aged peers were involved in the project only 23

featured in the complete sets of data for the 24 toddler participants. (One preschool-

aged peer featured twice in the dyad-play condition.)

The age of each peer in dyad play was to be maintained at 48 or more months.

Children of this age are normally skilled in language (this was confirmed by use of the

PPVT-R) and, as suggested by many researchers including Westby (1991) and those

referred to by Rubin, Fein and Vandenberg (1983) - for example Konner (1992) are

likely to possess the social and cognitive skills necessary to engage in social pretend

play at a level substantially beyond that usually exhibited by a toddler. However,

exceptions were made for three older peers, one aged 45 months and two aged 47

months, as they were clearly the toddler participants' 'self-chosen' familiar playmates.

(The three older peers' scores on the PPVT-R confirmed their suitability as partners for

dyad play.) The age range then for older peers was 45 to 63 months (SD = 5.01).

5.5.3.1 The Peabody Picture Vocabulary Test - Revised (Form M)

As mentioned earlier the Peabody Picture Vocabulary Test - Revised Form M

(1981) - was administered as a precaution to ensure older peers chosen for the dyad

play situation did not have developmental problems. (While the older peers may have

had other developmental difficulties which were not screened for, PPVT-R assessment

indicates they probably have the necessary language and intellectual functioning to

engage in age-appropriate symbolic play.)

The dyad-play condition required the participation of an older peer who was

well known to the toddler and where possible one with w h o m the toddler engaged in

play in the mixed-age play sessions. It is, however, inappropriate to include an older

peer who may have potential developmental difficulties or problems, as a collaborative

partner for the less-skilled toddler in dyad play. Receptive vocabulary is a reliable

indicator of verbal and cognitive development in preschool-aged children in Australia,

consequently assessment of the older peer's vocabulary skills was deemed as useful for

screening out children who may exhibit some developmental problems.

The Peabody Picture Vocabulary Test (PPVT) - Revised, Form M (1981) - was

considered the appropriate instrument for this purpose. This instrument, the PPVT-R

(Form M ) is widely employed by researchers and is regarded as suitable for use with

Australian participants (de Lacey, Walker, & Nurcombe, 1999). It is useful for

measuring language and vocabulary through identifying, for example, children's levels

140


of semantic comprehension and vocabulary proficiency (Dunn & Dunn, 1981). These

two components, together with the ability to make inferences, are essential skills for

engaging in effective symbolic play (Westby, 1991). Importantly, the PPVT-R also

correlates with scores on symbolic play (Wyver & Spence, 1999).

The PPVT-R (Form M ) was administered to the older peer before his or her

involvement in Condition Three (dyad play). Implementation involves the examiner (in

this case the present researcher) saying a word and the child responds by pointing to a

picture. R a w scores are calculated and converted to standard scores. These are "based

on age with a mean of 100 and a standard deviation of 15 - derived from a national

(USA) standardization sample" (Howes, 1997).

All older peers scored over 100 on the PPVT-R except two whose scores were

low, possibly due to English being their second language. The age range then for older

peers was 45 to 63 months (SD = 5.01). (A summary of the PPVT-R scores is in

Appendix D.)

5.5.3.2 Older-Peer Attendance at LDCCCs

A second prerequisite for the older peer was that they must have attended the

L D C C C for at least three months. (This requirement also ensured they had been

involved in the mixed-age play sessions for the same period.) A n exception was made

for one older peer who had attended the L D C C C for one month only (she did however,

attend every day) before observations commenced, as she too was clearly the preferred

play partner for the toddler involved.

Older peers favoured by the toddler participants during the mixed-age play

sessions were identified. Older peers were required for the dyad-play condition. If a

toddler indicated no clear preference for an older peer (as happened in four cases), then

staff were asked to suggest a child they deemed suitable for this role. Nine males and

fourteen females comprised the older peer group. As mentioned earlier, one female

peer was involved in two dyads as the 'chosen, familiar' older peer.

Characteristics of the older preschool-aged peer participants are presented in

Table 5.5.

141


Table 5.5

Characteristics of Preschool-Aged Participants

Participant Sex Age in N siblings, sex and age in N months enrolled in N days

Id months months LDCCC LDCCC

attended

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

*18

19

20

21

22

23

F

M

F

F

F

F

F

M

F

M

M

F

F

M

F

M

F

F

F

F

M

M

M

45

47

47

48

48

48

49

50

51

51

52

52

53

53

54

55

55

56

56

59

61

62

63

None

F18

None

None

Not known

None

None

M. 19 years

Not known

Not known

F31

M9

M33

None

Not known

F108

F 20 years

F92 F56

M32 F16

F23

None

F33

Not known

1

41

19

3+

3+

9

41

24

3+

48

3+

3

15

24

3+

48

3+

14

29

41

48

41

11

5

4

5

5

5

5

5

5

5

5

5

5

4

3

4

5

5

2

2

5

5

5

5

* denotes older peer who engaged with two different toddler participants in dyad play

142


Additional information relating to the older peer is displayed in Table 5.6.

Table 5.6

Summary of Older-Peer Participant Information

Age at time Period in L D C C C Weekly L D C C C of observation before data attendance

(months) collection (days) (months)

Mean 52.83 *20.65 4^52

Range 45 to 63 3 to 48 2 to 5

Standard 5.01 17.66 9.47 deviation

*this figure very approximate as for 6 participants the only information available for length of attendance in L D C C C was 'more than 3 months'

5.6 PROCEDURES FOR DATA COLLECTION

5.6.1 Implications Arising from Changes in Commonwealth Government Policy

During the initial recruitment period for the major study - undertaken in the

period February to April 1997 - many children in the targeted age group were attending

L D C C C s for more than 24 hours per week. However, the advent of reduced

government funding which occurred in July 1997 (after the recruitment process)

resulted in a substantial reduction in the number of possible participants. A secondary

effect was that some centres were forced to reduce staff, which also invariably affected

the daily program.

The withdrawal of children from L D C C C s or a reduction in the number of days

children attended greatly affected the remainder of the data collection. In time, several

other factors became evident as a result of the population movement experienced in

child-care centres. These factors included a change in the grouping of children within

the centre and rearrangement of timetables; in addition often the overall daily program

was adversely affected in terms of the researcher's requirements.

A further outcome of these changes was the decision to modify the study and

reduce the number of targeted participants. Logistically this was essential as the

researcher found it necessary to make many extra visits to some L D C C C s to

143


accommodate the children's attendance patterns. Changes might also have impaired the

external validity of the study.

5.6.2 Data Collection

5.6.2.1 Equipment

The data-collecting process required the collection of nine, ten-minute video

taped observations of participants in three conditions. A compact video-camera (Sony

Hi-8) with a liquid crystal display (LCD) screen was employed to record the

observations. This model of camera was chosen in preference to models requiring the

operator to look through the viewfinder. The L C D screen component together with a

most effective zoom feature (48 x digital zoom) made the camera both easy and discrete

to use and also allowed for speedy and effective 'screening out' of events or personnel

not to be included in the observations. Despite efforts to film discretely, from time to

time it was evident that children knew they were being filmed.

5 6.2 2 Video-Taped Observations

Nine ten-minute video-taped observations were collected of children engaged in

free play. Three of these observations were of participants engaged in outdoor free play

with their same-age peers - Condition One. Three such observations were of the

participants in an outdoor free-play session where their older preschool-aged peers were

present - Condition Two. The final three observations were filmed when each

participant was involved in what (for the purpose of this project) was described as dyad

play - Condition Three. In this condition each toddler participant was paired with a

familiar 'chosen' older peer. The dyads were then presented, on each of the three

occasions, with different play materials likely to elicit pretend play. The writer supplied

these materials.

5.6.2.3 Length of Video-taped Observations

Determining the length of time necessary to adequately observe play is a vexed

issue. For example, Fein and Apfel (1979) observed each of their participants, for a

period of ten minutes, on one occasion only. In contrast, three, 15-minute observations

of children in three different conditions, each observation conducted a week apart, were

collected by Howes and Farver (1987). Lamb, Sternberg, Knuth, Hwang, and Broberg

(1994) found in their study of peer play and non-parental care experiences that, even

144


with children as young as 16 months, 30-minute observations were sufficient to obtain

reliable measures of the quality of peer play.

In determining the length of observation periods the precedents just mentioned

were taken into account as well as the age of the participants, the dependent variable

and the context and the purpose of the research.

Therefore it was decided to collect 30 minutes of data for each participant under

each of the three conditions, thus a total of 90 minutes of data were gathered for each

participant. Data for all conditions were collected in 10-minute episodes, each segment

reflecting activity observed on a different occasion. As 24 participants were involved in

this current study, a total of 36 hours of video-taped observation data was collected.

W h e n play sessions were filmed early in the morning, no observations were

undertaken until each participant had been present in the play area for at least twenty

minutes. This precaution ensured toddlers were settled in to the child-care day.

5.6.2.4 Order of Observations

Counterbalancing was implemented to ensure that participants were not

observed in either the mixed-age play sessions or the same-age play sessions in any

regular order. This was achieved by establishing a rotating order of participants for

video-taping at the start of observations. If targeted children were absent, a make-up

session was arranged to collect the missing observation. O n occasions children were

removed from the environment for routine care-giving tasks and hence not available for

the planned observation. W h e n this occurred the observations were completed at the

end of the session.

5.6.2.5 Conditions under which Participants were Observed and Video-Taped

Video-taped observations were collected of toddlers engaged in play under the

following three conditions: condition one - same-age play; condition two - mixed-age

play and condition three - dyad play. It was imperative that the same-age play session

took place in environments as nearly identical as possible (using the same materials and

equipment) as provided in the second condition - mixed-age play.

A detailed record of toys and materials available for play in these sessions was

maintained. Also these materials were ranked in groups according to the potential they

were judged to have for supporting and encouraging children's pretend play. (Further

detail is presented in Chapter Six (6.12.2.1) where the descriptive data are analysed.)

145


A further consideration for both the same-age and mixed-age conditions was the

time of day these play sessions took place. In each participating L D C C C both same-age

play and mixed-age play happened twice a day. Mixed-age play usually occurred early

in the morning and towards the end of the afternoon. Same-age play generally took

place after the morning mixed-age play period and before the afternoon mixed-age

session

- Condition One: Same-Age Play

In order to fulfil the requirements of the first condition, toddlers were

required to play freely in the outdoor environment with a group of same-age

peers. Some variation occurred among the age-grouping methods employed by

the L D C C C staff. Nevertheless, in same-age play no child in the group was

more than 18 months older than the participant.

Three segments of video-taped activity, each ten minutes long, were

collected for each toddler participant under Condition One. Filming of toddlers

in Condition One and T w o (mixed-age play) was conducted first. As with the

mixed-age play observations, if same-age play sessions were held at different

times of the day every effort was made to observe children at each of these

times.

- Condition Two: Mixed-Age Play

The second condition required the toddler to participate in mixed-age

play. To fulfil this requirement toddlers were observed playing freely in the

outdoor environment in a mixed-age group that included the preschool-aged

children and may also have included children younger than the target toddler

group. While some variation existed among centres, generally these sessions

featured children aged approximately 16 to 18 months to more than five but less

than six years old sharing the same space and care-giving staff. There were, in

mixed-age sessions, approximately twice as many preschool-aged peers as there

were toddlers. The number of older peers the toddlers interacted or played with

in the mixed-age sessions ranged from one to ten - the average being 4.24

(Researcher's field notes). However, the data are not available to employ an

analysis, for example, as used by Goldman (1980) - w h o in any case found that,

in mixed-age groups, choice of play partner was not greatly influenced by age.

146


As is recorded in the field notes, in most centres a variety of materials

and equipment was available to cater for the differences in the children's

developmental levels.

Each ten-minute session of mixed-age play was filmed on a different

day. As mixed-age play sessions took place at different times of the day, for

example, in the morning and afternoon, every effort possible was made to

observe a range of morning and afternoon sessions.

- Condition Three: Dyad Play

The third and final condition, which required the toddler and an older

peer to engage in free play with materials supplied by the researcher, was

implemented towards the end of the data-collection period. Delaying

implementation to this point in the observation process allowed identification in

mixed-age play, in collaboration with staff when required, of an older peer who

appeared to be favoured by the toddler participant. (This timing also allowed the

participants to become more familiar with the researcher's presence in the

centre.) For the reasons stated earlier the dyad play condition required that an

older peer be aged approximately between 48 and 60 months to partner the

toddler. To fulfil supervision requirements an area aside from, but remaining

part of, the main indoor play area was used and the dyads presented with three

different sets of play materials on three separate occasions.

Before video-taping of the dyad situation began the older peer was

requested to, "Please help (name of child) to play with these toys". The

objective was to film each dyad-play session for ten minutes. However,

toddlers, especially younger toddlers, are keen to explore the environment and it

was often impossible to achieve a ten-minute video-taped observation. To this

end film segments of dyad play vary in length from two minutes to the full ten

minutes ideally sought. W h e n toddlers indicated they wished to leave the

situation before the ten minutes was completed, filming finished with their

leaving.

147


5.6.2.6 Provisions for Play in Condition O n e (Same-Age Play) and Condition T w o

(Mixed-Age Play)

The L D C C C staff planned and provided the toys, materials and equipment

available in the outdoor free-play environment for same-age play and mixed-age play

(Condition One and Condition Two). These provisions reflected the program planned

for the children by the centre staff. This researcher had perforce no influence upon what

was provided for these sessions.

5.6.2.7 Provisions for Play in Condition Three (Dyad Play)

T w o 'ranges' of play materials were provided by the current researcher for dyad

play (Condition Three). The first range consisted of toys and props that are familiar and

highly realistic. Materials of this nature support toddlers' emerging representational

skills. A plastic tea-set with cutlery and a coffee pot together with dolls comprised one

set, and another of play-dough with appropriate accessories (bowls, measuring cups and

spoons) completed this requirement. Similar play materials have been successfully used

by other researchers (Westby, 1991; Howe, Moller, & Chambers, 1994; Rusher, Cross,

& Ware, 1995).

The second range of play materials included representational, familiar and

unfamiliar play materials and toys, which require children to use "role and situational

transformations" (Westby, 1991, p. 147). This group of materials includes large blocks,

table blocks, Lego and constructional toys with small representational toys, figures and

vehicles similar to those used by Westby (1991) and Howes and Unger (1992). Table

blocks, wheeled vehicles and human-like figurines were supplied for this category.

Materials presented in Condition Three are detailed in Table 5.7.

Presentation of these play materials was always in the same order and in as

similar manner as was possible. As mentioned earlier each older peer was instructed to,

"Please help (name of child) to play with these toys". (See Appendix Fi.)

148


Table 5.7

Play Materials Supplied for Dyad Play (Condition Three)

Order in which play

materials were

presented

Description of play

materials

Type of symbolic play

likely to be elicited by use

of play materials

1 plastic tea-set, cutlery set, coffee

pot, two dolls

emerging representational

skills

table blocks, small, plastic,

wheeled-vehicles (two cars, two

trucks, two aeroplanes) two

human-like figurines

"role and situational

transformations" (Westby,

1991,p.l47).

play dough and accessories

(bowls, dishes, measuring cups,

measuring spoons, ice-block

containers)

emerging representational

skills

5.6.2.8 S u m m a r y of Data Collection Completed in each L D C C C

The period of time during which data collection occurred ranged from three

weeks to seven weeks depending on several factors. These factors included: the

participating children's attendance; spontaneous program changes in the L D C C C ; the

researcher's availability and inclement weather. Changes brought about by adjustments

in funding also had a negative impact upon the observation schedule as children's days

of attendance changed at very short notice and at times children were abruptly removed

from the L D C C C . A n outbreak of chicken-pox in metropolitan Sydney also extended

observation schedules in two centres.

These factors hindered data collection with the result that incomplete data

collections ensued for half the participants. A s can be seen in Table 5.8 although 48

toddlers participated in the project it was possible to complete only a total of 24 sets of

data.

149


Table 5.8

Summary of Total Data Collection

LDCCC Id

1

2

3

4

5

6

7

8

9

10

11

12

Category of

LDCCC


Children's services (community)

Municipal

Community

Private

Private

Municipal

Community

Community

Private


Community/ work-based

N participating

toddlers

4

8

4

2

5

5

5

1

5

4

3

6

N sets of data

obtained

0

0

4

2

2

3

2

0

5

0

2

4

It is important to note that the L D C C C s where incomplete sets of data were

obtained are those where observations were undertaken around the time the funding

policy was changed. The incomplete sets of data reflect the activity and fluctuations in

children's attendance patterns as parents tried to come to grips with the increase in fees

and what many viewed as a reduction in quality of services as staff were forfeited.

In conclusion, as presented below in Table 5.9, a complete video-taped data

collection for each participant comprised the following:

- three, ten-minute segments of the toddler engaged in same-age

play (Condition One);

150


three, ten-minute segments of the toddler engaged in mixed-age

play (Condition Two);

three, ten-minute segments of the toddler engaged in dyad

play (Condition Three).

Table 5.9

Summary of Complete Data Collection

Experimental

Group

Group A

Males with

older sibling

GroupB

females with

older sibling

Group C

males without

an older

sibling

Group D

females

without an

older sibling

N

participants

in group

5

5

6

8

N 10-minute

observations

of same-age

play

(Condition

One)

3

3

3

3

N 10-minute

observations

of mixed-age

play

(Condition

Two)

3

3

3

3

N 10-minute

observations

of dyad play

(Condition

Three)

3

3

3

3

Total

observations

in minutes

90

90*

90**

90*

*Indicates four toddler participants who stayed less than 20 minutes in Condition Three

(dyad play).

151


5.7 INSTRUMENTS FOR ADDITIONAL DATA COLLECTION The following instruments, each discussed in detail below, were used to collate

descriptive data seen as essential to the project:

- Environment Rating Scale;

- Staff questionnaire (including details of training and qualifications in

Early-Childhood Education);

- Parent questionnaire.

Information arising from the use of these instruments in conjunction with staff-

ratios and materials provided for play was explored in association with participants'

symbolic-play behaviour. (Further discussion is found in later in this chapter in 5.9.6.)

5.7.1 Parent Questionnaire

A questionnaire was constructed for completion by parents of the participants

(see Appendix Civ). In an endeavour to ensure the questionnaire was likely to be

effective in terms of the information required, it was submitted for comment to several

established Australian developmental researchers. The document was modified to

include any suggestions, put forward by those reviewers, likely to increase the

usefulness of the questionnaire.

The instrument was designed to primarily elicit and gather information, relating

to family characteristics and social context, considered relevant to the project. Parents

were asked for the following information:

- the reason for enrolling their child in L D C C ;

- the reasons for their choice of L D C C C ;

- their expectations in terms of their child's development as a result of

attending the L D C C C ;

- the language predominantly spoken at home;

- the age and sex of other children in the family;

- typical activities engaged in by the participating child at the

weekend;

- the ages of each parent;

- the level of education reached by each parent and their current

occupation.

152


Parents were requested to complete this questionnaire and return it to the

researcher. The response rate was pleasing with 91.5 percent of the questionnaires

completed.

5.7.2 The LDCCC Environment

The early-childhood literature indicates that a degree of consensus exists among

researchers in terms of the degree to which mixed-age free play benefits younger

children. A n anomaly remains, though; as to the degree this impact is moderated by the

quality of care-giving reflected in the long-day child-care environment. Quality with

respect to the various aspects of the physical and caring environment has yet to be

considered as a substantial independent variable in a study focusing on mixed-age free

play.

Components of the care-giving environment include: learning activities such as

the provision for symbolic play and active physical play, staff-child interactions and

discipline or behaviour management, and personal care routines such as toileting and

mealtimes. Scarr, Eisenberg, and Deater-Deckard (1994) identified these features of the

L D C C programme as 'process quality indicators'. Research conducted by Howes,

Phillips, and Whitebook (1992) suggests that process quality aspects of the setting (such

as these included in the Environment Rating Scale) have a stronger relationship with

child outcomes than the structural elements (group sizes, staff training) of the care

giving environment. However, it is clear that structural elements may influence the

process variables.

5.7.2.1 The Environment Rating Scale

A n Environment Rating Scale was used to further enhance the description of the

care-giving environment. The Infant and Toddler Environment Rating Scale - ITERS -

(Harms, Cryer, & Clifford, 1990) is a rating scale frequently used by other researchers

including: Howes et al. (1992) and Dunn (1993). However, the findings of a study by

Scarr et al. caused them to postulate, "all aspects of quality are so highly correlated that

any set of theoretically-chosen items will capture not only the selected aspect of quality

but all others as well" (1994, p. 147).

Scarr's research team undertook to evaluate how effectively three widely used

assessment scales one of which was the 35 item ITERS, measured the quality of child

care. Each scale contained items, which measured L D C C quality in terms of both the

153


regulatable and process components. Regulatable items include group sizes, ratios of

adults to children and salary scales. Process quality indicators include, for example,

those that assess interactions between staff and children or the adult's support of the

child's developing language skills. Scarr et al. (1994) concluded that extensive diversity

in child-care quality existed especially in terms of regulatable items - this difference

arising from the range of governing bodies and the variety of regulations implemented

by these institutions.

The findings arising from their study also highlighted the high intercorrelation

that existed among process indicators and the provision of quality care. Consequently,

Scarr et al. (1994) concluded that it was appropriate, especially for research purposes,

when efficiency in terms of cost and time is at a premium, for investigators to employ a

shorter scale focusing on the process aspects of the environment. The decision, to

construct a scale featuring a reduced range of items, was based on these conclusions.

To that end an Environment Rating Scale was constructed - adapted from the

Infant and Toddler Environment Rating Scale (Harms et al., 1990). Twelve items were

compiled in accordance with the findings of Scarr et al. (1994).

The current study was undertaken in L D C C C s in the state of N e w South Wales,

Australia where regulations governing such institutions are consistent across the state.

Such conditions supported the researcher's decision to compile an abbreviated

environment assessment scale of twelve items composed largely of process aspects of

the day-care environment (see Appendix E).

T w o Environment Rating Scales were completed for each centre: one,

independently by this researcher, and the second by the centre Director. Each of these

assessments was undertaken at the end of the data-collection period. This timing

supported this writer's belief that, first, her o w n knowledge of the centre would by then

be extensive and allow a more precise rating to occur and, secondly, that the Director

would be more at ease with both the researcher and the request to complete the scale.

The first Environment Rating Scale was completed independently shortly before

the final observations were made. The second assessment was completed by the

L D C C C Director in the company of this writer at the conclusion of the observation

period. Care was taken not to influence the Director's ratings. The information

recorded in these scales further supported the description of the child-care environment.

Importantly the Environment Rating Scale primarily provides a gauge for

assessing the day-to-day quality of the care-giving environment. A Likert scale was

154


adopted in accordance with Harms et al. (1990). A rating of 1 on the Likert scale (of 1

to 7) indicates an inadequate provision of the aspect of care under assessment - while 7,

the highest rating, indicates excellence. As there were 12 items each with a possible

score of 7, the m a x i m u m rating achievable was 84. A score of 45 or less indicated the

existence of a minimal - unsatisfactory - provision of care. Scores between 46 and 50

were recognized as indicating the provision of care-giving was 'adequate'. These

scores indicated that the provision of care was above 'minimal' or unsatisfactory care,

but below that categorised as 'good' quality care. L D C C C s where the score was 50 or

above were considered to be providing 'good' or 'better than good' quality care. N o

centre had a rating below 3 (which was synonymous with minimal care) on any item.

The average Environment Rating Scale score over the eight centres used in the research

was 52.5 suggesting that overall the quality of the care-giving environment was good.

5.7.2.2 Toys, Materials and Equipment

The form and content of children's play depends in part upon the availability of

appropriate play materials. Developmentally and culturally appropriate toys, materials

and equipment support the child's "retrieval of play schemas" (Westby, 1991, p. 147).

Symbolic or pretend play is essentially a creative activity and thus requires a range of

play materials to support the development and complexity of such play.

In Australia, owing to favourable climatic conditions, provisions for outdoor

play in early-childhood settings lend themselves to replicating much of what would, in

other climates, be viewed as indoor activities. In addition to equipment to promote and

support children's gross motor development, toys, materials and equipment, the use of

which enhance children's skills in all areas of development, are to be observed in the

outdoor environments of L D C C C s .

A n inventory of most outdoor play areas in L D C C C s would reveal provisions

for symbolic play in the following categories: realistic props to support children's

engagement in familiar everyday activities in which they themselves participate and

experience; props to support home or domestic play eliciting play which replicates that

of familiar others; small replica play materials to encourage children's engagement in

both familiar and less familiar symbolic-play activities; construction equipment such as

unit blocks, mobilo or Lego; and an assortment of open-ended props such as rugs, cloths

and cardboard boxes to be used as children require in their imaginative symbolic play

155


(Rubin, Fein, & Vandenberg 1983; Westby, 1980, 1991; Howes & Unger, 1992;

Gowen, 1995). (See Appendix Fii.)

Detailed records of the toys, materials and equipment during all video-taped

observations were maintained. Play materials likely to support symbolic play were

grouped in accordance with the categories suggested by Westby (1991) namely, high-

realism toys, small representational toys (ideally these would be in the vicinity of

construction toys to encourage the combination of construction and pretend-play

activities), realistic props likely to support thematic play, and gross-motor equipment.

It is important to note that, while Westby (1991) identified her suggested toys,

materials and equipment as suitable for indoor play spaces, they are all considered

appropriate for outdoor play in the L D C C C s that participated in the current study,

conducted in Sydney, Australia. Here, the climate allows for a replication of indoor

play activities in the outdoor environment at any time of the year. While the hours

allocated for outdoor play vary from summer to winter (in response to the sun's

movements and strength), provisions for play remain much the same.

5.7.2.3 Ratios of Staff to Children

As discussed earlier in this chapter (5.7.2.1) an Environment Rating Scale to rate

the 'process' quality indicators, was completed for each participating L D C C C . This

Environment Rating Scale, based on the findings and recommendations of Scarr et al.

(1994), did not assess 'structural' components, one of which is the staff-child ratio, of

the care-giving environment. Ratios of staff to children are an important component of

L D C C C quality and this information was therefore deemed as essential in further

describing the participants' play environments.

Minimal staff-child ratios in the centres where data were collected for this

project must conform to N e w South Wales state regulations. These regulations stipulate

that one adult is required for each group of five children under two years old; one adult

is required for each group of eight children aged two to three years old while one staff

member is required for each group of ten children aged four to five or more years old.

In the current project, data on staff-child ratios were collated by the following

method. The numbers of adults and children in the outdoor environment at the

beginning and conclusion of each observation session were counted and those numbers

recorded. These counts were approximate and numbers of both staff and children

tended to fluctuate during the video-taped sessions.

156


5.7.2.4 Staff Questionnaire

A nine-item staff questionnaire was devised (see Appendix Bv). The questions

were developed to identify staff beliefs and values relating to influences on children's

development. A Likert scale was constructed for the responses to the seven items

relating to children's well-being. The remaining two questions required staff first, to

identify what they considered to be indicative of children's engagement in pretend play

and secondly, to list what they thought to be the main influences upon children's play.

O n the reverse side of the questionnaire a list of specialised early-childhood

qualifications was displayed (see Appendix Bvi). Staff were asked to indicate the level

of specialised qualification in Early-Childhood Education they may have obtained. The

qualification categories were then weighted in accordance with a scale used by Howes

and Smith (1995). Responses to these questionnaires were confidential and names were

not required.

The responses, to the following four statements only, were collated for analysis

and discussion:

- Item 5 - 'Please describe what indicates to you that children

are engaged in pretend play'.

- Item 6 - 'Please list what you consider to be five major

influences on children's play'.

- Item 7 - 'Play positively influences children's cognitive

development'.

- Item 8 - 'All children benefit from taking part in mixed-age

free-play sessions'.

(Space was provided for respondents to answer Items 5 and 6. Staff answered

Items 7 and 8 on a five-point Likert scale, 1 = strongly disagree to 5 = strongly agree.)

A meeting with LDCCC centre staff followed the initial contact with each

participating centre, and the format of the questionnaire was discussed at this meeting as

part of the orientation process for project participants. Questionnaires, accompanied by

a letter thanking staff for their participation in the study (see Appendix Biv), were

distributed approximately halfway through the data-collection period and were returned

to a collection box at the L D C C C . At the end of the data-collection period

questionnaires were redistributed for any member of staff w h o m a y have missed the

first distribution. Stamped, addressed envelopes were supplied at this stage for the

157


return of the documents. A third and final distribution was made through the Director in

centres where the initial response had been very low accompanied by the suggestion

that, if the respondent preferred, an interview or telephone call could be arranged in lieu

of completing the questionnaire. The response rate of 70 percent was very satisfying -

of the 83 questionnaires distributed 58 were returned.

5.8 CODING AND SCORING PROCEDURE

Nine, ten-minute video-taped observations of participants' play were collected -

three segments in each of the three Conditions: same-age play, mixed-age play and dyad

play. Observations of participants in Condition One and Condition T w o were

undertaken in exactly the same outdoor environment where play materials were those

supplied by the L D C C C staff. Observations for Condition Three took place in the

indoor activity area with materials supplied by the researcher.

Each ten-minute video-taped observation was coded every 30 seconds for level

of symbolic play - the focus of investigation in this project. Data were also coded for

level of play (in terms of cognitive requirement) and thematic unit of play.

T w o coding sheets were used in conjunction with one another for scoring

symbolic play (see Appendices Gi, Gii). The first sheet enabled the coding of symbolic

play, level of cognitive play and thematic unit of play. This sheet also accommodated

the following information: participant details (including: name, sex, age); date, time and

site ( L D C C C ) of observation; whether Conditions One, T w o or Three; and the numbers

of staff and children present in the play environment. T o provide additional social

contextual information, the presence of peers or siblings, whether the same-age,

younger or older than the participant, was also noted on this initial sheet. Other

anecdotal information, perceived by the researcher, at the time of viewing the video

tapes, as likely to be useful was also noted on sheet one.

The second sheet provided additional space allowing the finer details of

symbolic play (dimension and level) to be recorded in accordance with Westby's (1991)

Scale for Assessing Children's Pretend Play.

158


5.8.1 Inter-Observer Agreement

T w o trained raters were involved in the coding reliability process. Each rater

was blind to the other and each had a television screen. The screens were placed in such

a way that the raters were facing opposing directions. The present author controlled the

video-recorder and stopped the tape every 30 seconds to allow rating to occur.

5.9 CODING FOR PLAY

5.9.1 Symbolic Play

Episodes were examined and coded for symbolic play using the Westby (1991)

Scale for Assessing Children's Pretend Play (see Appendix Ai). The use of the Westby

(1991) instrument enabled the identification of changes in children's emerging and

developing symbolic play whose four behaviours (or dimensions) she identified as


relationships' - as noted in Chapter One (1.4.1). Function and form of language are

also incorporated in this scale, but being irrelevant to this study were not used. While

Westby's (1991) scheme details eight different developmental levels of symbolic play

activity, only the first four levels were seen as appropriate for the this study.

The scoring of symbolic play was collated in three different ways. First, the

frequency of symbolic-play events was scored. This score resulted from ignoring

dimensions and levels of symbolic play and scoring all 30-second periods in which the

participant exhibited any symbolic-play activity. The m a x i m u m frequency achievable

was 20 in any one observed period - a total of 60 per condition. For the toddler

participants who did not play for ten minutes - those w h o truncated the dyad-play

sessions - the observed frequencies were weighted according to the length of stay. This

assumes that the frequency of symbolic-play events would have continued at the same

rate as observed. Secondly, the level of symbolic-play activity, in each dimension of

symbolic-play behaviour ('decontextualisation', 'thematic content', 'organisation of

themes' and 'self-other relationships'), for each 30-second segment of video-taped

observation was scored. Finally, the most-frequently occurring level of symbolic play

in each 30-second period was scored. (The above information, initially collated on the

two separate scoring sheets, was then entered on an 'individual summary sheet' - see

Appendix Giii.)

159


All components of the first four symbolic levels (Levels One to Four) identified

in Westby's (1991) assessment scale were appropriate for use in this study. However,

the 'decontextualisation' component of level six was included in the event of usage of a

replica toy occurring. Isolated instances of symbolic-play activity falling outside the

range of Levels One to Four were also coded and scored.

In summary, toddlers performing symbolically at Level One (17 to 19 months)

of Westby's instrument are likely to use life-like props in self-representational pretend

play involving single pretend actions and reflecting familiar everyday events.

Participants displaying the features of Level T w o symbolic play may be

expected to incorporate life-like props with behaviours indicating activities of familiar

others. They are likely to combine two actions and possibly direct these towards other

objects or people (as opposed to themselves as in Level One).

Progress to the third level of symbolic play requires children, while maintaining

the use of life-like props, to reflect the activities of both familiar others and known

events in their pretend. They may now direct this pretend activity toward objects or

people and engage in more elaborate schemas or scripts depicting the emergence of

simple sequences.

Finally the fourth level allows the observer to possibly identify the enactment of

experiences less-frequently encountered by the child represented by a more complex

sequence of actions.

The fifth level of Westby's (1991) assessment scale, when children are likely to

re-enact an experience or an event, as mentioned before, was thought to be outside the

range of expected symbolic-play behaviours in this study. Other features of the fifth

level include lengthier and more complex script use, the likelihood of the pretend player

transforming him or her into a role and engaging in associative play.

It was predicted and subsequently observed that, when features of the fourth and

fifth levels of symbolic activity emerged, which according to Westby's continuum is

when children are approximately 30 months old, children's symbolic play was soundly

supported and maintained by the use of language.

5.9.2 Coding for Level of Cognitive Effort

This was to be coded as described in Chapter Four, 4.8. However, no evidence

of level of cognitive effort was discernible in the data. Consequently this measure was

discarded.

160


5.9.3 Coding for Length of Thematic Unit

The length of thematic unit was to be coded in accordance with the description

in Chapter Four, 4.9. At no time during the observations was the children's symbolic

play activity sustained long enough to employ this measure. Therefore this measure

was also discarded.

5.10 ADDITIONAL INFORMATION

5.10.1 Social Play

Social play was noted in accordance with Parten's (1932) six categories -

unoccupied, onlooker, solitary, parallel, associative and cooperative play. (These were

described in detail in Chapter Four, 4.6.)

5.10.2 Cognitive Play

Cognitive play was recorded as nonplay, exploratory (or sensori-motor),

physical, practice (or functional) and constructive. (Each of these play categories is

described in detail in Chapter Four, 4.7.)

5.10.3 Imitation

The findings of Mandler and McDonough (1995) and other researchers

((Tomasello et al., 1993; Hanna & Meltzoff, 1993) support the usefulness of imitation in

acquiring skills and knowledge and suggest that the inclusion of a measure of imitation

is both useful and informative. Imitation which may be either verbal or gestural (action)

was noted as either present or not present.

5.10.4 Older-Peer Communicative Behaviour

To provide additional social contextual information, the presence of peers or

siblings, whether the same-age, younger or older than the participant, was also noted.

Older-peer communicative behaviour in Condition Two and Condition Three was

recorded. Both toddler participant imitation and older peer communication were noted

on the first coding sheet while the third coding sheet accommodated a more detailed

recording of the older peers' communicative behaviour (see Appendix Gv).

161


Imitation and older-peer communicative behaviours were recorded and noted by

the observer only - no inter-observer ratings were carried out. Hence neither of these

variables or influences were included in the statistical analysis.

5.11 ANALYSIS OF DATA

5.11.1 Choice of Statistical Tests

It was decided, that because the dependent variable was not normally distributed

and the number of participants was small, to adopt a conservative approach to statistical

analyses and use non-parametric tests. Inferential data are likely to arise from this study

in which the dependent variable is not normally distributed. Consequently, while

accepting that they may not be as powerful as parametrical statistical tests, non-

parametric tests will be used to analyse these data.

Four statistical tests will be employed to analyse the data: Friedman's Rank Test

(a two-way analysis of variance by ranks); The Kruskal-Wallis Test (a one-way analysis

of variance by ranks); Wilcoxon's Matched-Pairs Signed-Ranks Test; and Logistic

Regression Analysis (a method of computing 'odds ratios' or 'likelihood'). The sample

in this project was small - 24 participants only - hence the investigator adopted a

conservative approach when executing statistical analysis of the data.

5.11.2 Computer Software Packages

T w o software packages were used for the statistical analysis - M r N T T A B (The

Student Edition of Minitab for Windows and Minitab 12) and S A S (Release 6.12). The

former, M T N I T A B , featured three of the statistical tests used - the Friedman Two-Way

Analysis of Variance by Ranks, the Kruskal-Wallis One-Way Analysis of Variance by

Ranks and the Wilcoxon's Matched-Pairs Signed-Ranks test. The program S A S

featured Logistic Regression Analysis - a fourth statistical test required for data

analysis.

Variables influential on the participants' symbolic-play activity will be

identified using Logistic Regression Analysis.

5.11.3 The Friedman Two-Way Analysis of Variance by Ranks

The highest scores of symbolic play achieved by each participant were

compared for the three conditions (same-age play, mixed-age play and dyad play) using

the Friedman two-way analysis of variance by ranks - a non-parametric test (Siegel,

162


1956, p. 166-173). 'Child', that is each individual participant, was the blocking factor,

the second classification, was 'condition' (Siegel, 1956). These analyses were carried

out separately for the four dimensions of symbolic play: Dimension One,

'decontextualisation'; Dimension Two, 'thematic content'; Dimension Three,

'organisation of themes'; and Dimension Four, 'self-other relationships'. The test then

determines whether the rank totals differ significantly.

In summary the Friedman test illustrates "whether the size of the scores depends

on the conditions under which they were yielded" (Siegel, 1956, p. 173). (An extract of

the output from the Friedman's Rank test is in Appendix Hi.)

Similar analyses were carried out using the variable 'number of events of

symbolic play' exhibited during the nine observation periods (each condition was

observed three times). A Friedman's two-way analysis of variance was also carried out

for the variable 'most-frequently occurring level of symbolic play' exhibited by

participants during the three conditions.

Where the overall comparison of the three conditions - same-age play, mixed-

age play and dyad play - was statistically significant, further detailed examination was

carried out using the Wilcoxon's Matched-Pairs Signed-Ranks test (Siegel, 1956, p. 75-

83).

Use of the Friedman's Rank test enabled the current researcher to ascertain

whether the data supported the major hypothesis: that toddlers engaged in free play with

an older peer (in mixed-age play and dyad play) will exhibit symbolic-play more

frequently and at a higher level, than when they are engaged in free play with their

same-age peers. These Friedman's analyses in conjunction with the Wilcoxon's

Matched-Pairs Signed-Ranks test allowed the researcher to respond to the first

additional research question: if the major hypothesis was supported, which setting (or

condition) - same-age play, mixed-age play or dyad play - is more influential on the

toddlers engagement in symbolic play?

Results arising from the use of the Friedman test are presented in Chapter Six,

6.6,6.7 and 6.8.

5.11.4 Wilcoxon's Matched-Pairs Signed-Ranks Test

The magnitude and direction of the results from Friedman's Rank Test (as

related samples) were tested using Wilcoxon's Matched-Pairs Signed-Ranks Test

(Siegel, 1956, p. 75-83). Results are presented in Chapter Six, 6.6 and 6.7.

163


5.11.5 The Kruskal-Wallis O n e - W a y Analysis of Variance by R a n k

To examine the relationship between birth order and age group on participants'

symbolic play, the Kruskal-Wallis analysis of variance by ranks was carried out. The

Kruskal-Wallis tests "whether k independent samples could have been drawn from the

same continuous population" (Siegel, 1956, p. 193). Analysis was carried out

separately for each condition (same-age play, mixed-age play and dyad play) and each

dimension of symbolic play ('decontextualisation', 'thematic content', 'organisation of

themes' and 'self-other relationships').

It was then possible to respond to the following research questions: whether or

not first-born participants exhibited symbolic-play skills more frequently than later-bom

participants; whether or not first-born participants exhibited higher level symbolic-play

skills than later-born participants; and whether or not a difference existed between the

most-frequently occurring levels of symbolic play engaged in by first-born participants

and later-born participants.

The Kruskal-Wallis test was also used to ascertain whether the frequency of

symbolic play, the highest levels of symbolic play and the most-frequently occurring

levels of symbolic activity differed according to the toddlers' age group. Results are

presented in Chapter Six, 6.6.1, 6.6.2, 6.7.1, 6.7.2, 6.8.1 and 6.8.2. (An extract from the

Kruskal-Wallis output is in Appendix Hii.)

5.11.6 Logistic Regression Analysis

It was decided to examine the data in terms of whether the levels of symbolic

play engaged in by the participants were in one of two classes - either 'greater than' that

expected for their age or 'less than or equal to' that expected. That is, either more, or

less complex than expected in accordance with Westby's (1991) Scale for Assessing

Children's Pretend Play. This binary variable, which was named 'complexity', was

found for each child under each of the three conditions (same-age play, mixed-age play

and dyad play) and each dimension of symbolic play ('decontextualisation', 'thematic

content', 'organisation of themes' and self-other relationships').

The two classes are, 1 = 'greater than' and 0 = 'less than or equal to'. Such an

investigation required a Logistic Regression Analysis - a method akin to an analysis of

variance (Mc Neil, 1996).

Logistic Regression Analyses were carried out to examine the influences on

'complexity' of the following: the three conditions (same-age play, mixed-age play and

164


dyad play; the provision of play materials; the rating awarded the L D C C C on the

Environmental Rating Scale; the level of education achieved by the participants'

parents; and the participants' weekly attendance at L D C C C . Discussion of results and

accompanying tables are in Chapter Six, 6.9. (See Appendices Hiii and Hiv for an

explanation of the use of Logistic Regression Analysis and an extract from the Logistic

Regression Analysis output.)

5.12 ANTICIPATED RESULTS

In view of the research reviewed and initial observations, it appeared likely that

the participants will, first, readily engage in symbolic play with a more proficient older

peer during mixed-aged play sessions. It seemed likely that this joint or collaborative

play would occur in both a group mixed-age play setting (Condition T w o ) and in dyad

play (Condition Three). Secondly, it seemed likely that, when engaged in play with an

older peer in either Condition T w o or Condition Three, the toddler participant would

exhibit symbolic-play skills more frequently than when not engaged with older peers. It

also appeared likely that the symbolic play engaged in by the toddler with a more

skilled and competent other, namely an older peer, might be more sophisticated and

complex than if he or she were interacting with a same-age peer.

The context of a mixed-aged group engaged in free play maximises the

opportunity for the older peer to provide information and for the younger peer to

practise new or emerging skills in a manner subsumed by the Vygotskian notion of the

'zone of proximal development'. This same context allows the younger child to

participate in more complex activities than if he or she was playing alone or with

another child at the same stage of development.

In terms of Piagetian cognitive theory, this context provides the stimulus for the

toddler to experience the conflict necessary to promote the restructuring and redefinition

of schemes - the process of assimilation.

5.13 CONCLUSION

The methodology and analysis having been explained, w e now turn to a

presentation of the results in the next chapter.

165

Chapter 6: Results

CHAPTER 6: RESULTS

6.1 SUMMARY OF DATA COLLECTION PROCEDURE

As explained in detail in Chapter Five, data were collected in nine L D C C C s in

the inner-city area of Sydney. Twenty-four participants, thirteen females and eleven

males, aged from approximately 18 to 30 months, were involved in this project.

Approximately half the participants had an older sibling. Each toddler was video-taped

under three conditions: a same-age play session in the outdoor play environment

(Condition One); a mixed-age play session with preschool-aged peers in as close as

possible an environment as that shared with their same-age peers (Condition Two) and

finally, dyad play where the toddler was paired with a familiar older preschool-age peer

(Condition Three). The research method required each participant to be observed for

ten minutes in each condition, on three different occasions.

Materials for the first two conditions were those generally available at each

L D C C C while the researcher supplied materials for the dyad play. To ensure

comparability a record was maintained of the range and type of materials available

during the mixed-age play and same-age play sessions.

As, in dyad play, the older peer was asked to assist the younger child to play

with the materials provided, it was important that the older peer's development was

within the range expected for preschool-aged children. T o this end, the Peabody Picture

Vocabulary Test Revised - Form M was administered to the older peer before the dyad

condition was implemented.

A range of other information was gathered to supplement that obtained through

the video-taped observations, including whether or not the participant had an older

sibling and the number of days (three, four or five) the participant attended the L D C C C .

The writer, both independently and in conjunction with the Director of the

L D C C C , completed an Environment Rating Scale - adapted from the Infant/Toddler

Environment Rating Scale devised by Harms, Cryer, and Clifford (1990).

A Staff Questionnaire, designed to elicit information concerning attitudes to a

number of features of the L D C C C program including those towards play, was

distributed and collected. The questionnaire also allowed for staff to identify their level

of training in Early-Childhood Education. (Seventy-three staff questionnaires were

distributed and 58 were returned.)

166

Chapter 6: Results

Parents of the toddler participants completed the Parent Questionnaire, which

identified parental education, employment, expectations for their child's development

while attending day-care and examples of family leisure time activities. Of the 24

questionnaires distributed, 22 were returned. A summary of project components is

presented in Table 6.1.

Table 6.1

Summary of Project Components

Toddler Participant Information

Age range N female

18-30 months 13

Nmale

11

N days per week attended L D C C C

3, 4, or 5

Methodology - Observed Conditions

Condition Play materials supplied by:

Length of each observation

Total minutes observed

1.Same-age play Centre staff 10 30

2. Mixed-age play Centre staff 10 30

3. Dyad play Researcher 10 30

Measures Used

PPVT-R Environment rating Staff questionnaire Parent scale questionnaire

Toddler Quality of care Attitudes toward Level of education participants' provided by play, training and achieved, age,

receptive language L D C C C qualifications in employment, Early-Childhood developmental

Education expectations for child while in L D C C , home

language, family leisure activities.

167

Chapter 6: Results

6.2 CODING

As described in Chapter Five, the data (video-taped observations) were then

coded every 30 seconds for symbolic play using Westby's (1991) Scale for Assessing

Children's Pretend Play. Westby's scale was selected as it allows a very finely detailed

level of analysis of symbolic-play behaviours. The scheme accommodates the coding

of four symbolic-play behaviours or dimensions: 'decontextualisation', 'thematic

content', 'organisation of themes', and 'self-other relationships' over eight levels of

symbolic-play development as identified by Westby (1991). These levels encompassed

the expectations for symbolic play development for children aged from 17 to 60 months.

Each level indicates the changes in cognition required to bring about a modification in

symbolic-play behaviours. Westby also identified those dimensions that remained

stable while others changed. Of the eight levels featured on the scale, only Levels One

to Six were useful for this study. (Details of Westby's scale are in Appendix Ai.)

6.2.1 Scoring Symbolic Play

The coded video-taped observation data were scored in three ways. First, the

frequency (number of events) of symbolic play was scored for each of the three

conditions. This entailed tallying the number of 30-second periods showing symbolic-

play behaviour for each condition. (All levels of symbolic play were collapsed for this

process.) A maximum total of 20 symbolic-play 'events' (or instances) was possible in

each condition so a total of 180 events was possible if scores for the three conditions

were combined. Scoring was completed over all three conditions. A score of zero was

allocated where no symbolic play occurred.

Secondly, the highest level of symbolic play for each 30-second period of

symbolic play was scored in accordance with the Westby (1991) scale of assessment.

This score could range from one to six as, although the scale has eight levels, no child

scored above level six.

Finally, a score was given for the most-frequently occurring level of each

dimension of play behaviour ('decontextualisation', 'thematic content', 'organisation of

themes' and 'self-other relationships') shown over the total period of each condition.

As each 30 seconds of play was scored, the highest possible score for the most

frequently occurring symbolic-play behaviours was 60 for each of the three conditions:

same-age play, mixed-age play and dyad play. (Therefore the maximum score possible

if scores in all three conditions were combined was again 180.)

168

Chapter 6: Results

6.2.2 Inter-Observer Agreement

T w o inter-observers were involved in this process. One observer was a

supervisor, the other a graduate of a university. Both were trained in the use of the

coding system employed by this writer and each rated entirely independently. The first

viewed 8 percent of the total data - 3 percent of this data was randomly selected and 5

percent was specifically selected - enabling the discarding of long periods of inactivity.

The second rater viewed 4 percent of the total data - all randomly selected. Hence a

total of 12 percent of the data was coded by a second observer.

The amount of data subjected to inter-observer agreement appears to vary

considerably. In a review of six studies a second rater coded the following percentages

of total data: 100 percent (Haight & Miller, 1992); 20 percent (Bornstein, Haynes,

Legler, O'Reilly, & Painter, 1997); 12 percent (Bornstein, Haynes, O'Reilly, & Painter,

1996); 10 percent (Damast, Tamis-LeMonda, & Bornstein, 1996); 10 percent (Howes,

Hamilton, & Matheson, 1994); and 9 percent (Holloway & Reichhardt-Erickson, 1989).

In the light of these findings a goal of 12-15 percent of data to be coded by a second

person seemed reasonable.

In order to establish the level of inter-observer agreement the researcher opted to

use 'exact agreement' method for this project. This method, regarded as a more

stringent method than others, was also chosen as it is "more sensitive to agreement on

specific instances of behaviour than other indexes ..." (Page & Iwata, 1986, p. 110). As

stated, this system is conservative and sensitive and therefore may not result in

indicating reasonable levels of inter-observer agreement.

(An 'interval agreement' method could also have been used where the reliability

is the percentage of intervals in which the raters agreed. However, using this system

with data where behaviour occurs either often or rarely can result in misleading

information as some rating would be due to chance.)

As the experimental conditions were such that both high and low occurrences of

the target behaviour were observed the writer adopted the component 'Mean-

Occurrence - Non-occurrence Agreement' as suggested by Page and Iwata (1986, p.

115).

169

Chapter 6: Results

Therefore inter-observer agreement was calculated in two stages. First, the

inter-observer agreement for occurrence and non-occurrence of each of the four target

behaviours (decontextualisation, thematic content, organisation of themes and self-other

relationships) was computed using the following formula:

A freq x 100

A freq + D

(Poling, Methot, & LeSage, 1995, p. 74).

Secondly, the means of each of these two scores were then calculated and the

results reported as the indices of agreement. The 'A' frequency represented the number

of intervals for which the two observers recorded exactly the same number of

occurrences for the target behaviour and the 'D' letter the number of intervals in which

the two observers did not record exactly the same number of occurrences of the target

behaviour. T w o observers were involved in the process the first viewing 4 percent of

the data, the second 8 percent. Therefore a total of 12 percent of the data was subject to

inter-observer agreement.

A high agreement among observers was recorded. The results are presented in

Table 6.2.

170

Chapter 6: Results

Table 6.2

Inter-Observer Agreement

Observer Observer 1 2

% Agreement for occurrence of target behaviour 'Decontextualisation'

% Agreement for non-occurrence of target behaviour 'Decontextualisation'

% Agreement for occurrence of target behaviour 'Thematic content'

% Agreement for non-occurrence of target behaviour 'Thematic content'

% Agreement for occurrence of target behaviour 'Organisation of themes'

% Agreement for non-occurrence of target behaviour 'Organisation of themes'

% Agreement for occurrence of target behaviour 'Self-other relationships'

% Agreement for non-occurrence of target behaviour 'Self-other relationships'

100% 90%

86.71% 97.67%

82.08% 90%

89.20% 96.94%

75% 77.35%

88.92% 95.23%

83.63% 77.77%

88.53% 93.82%

% Mean of agreement

95%

92.19%

86.04%

93.07%

76.17%

92.07%

80.7%

91.17%

6.3 STATISTICAL ANALYSIS

As described earlier in Chapter Five (5.11) data analysis involved the use of four

major statistical tests. These were selected to enable a conservative analysis of the data.

These statistical procedures are non-parametric as the data are obtained from a relatively

small sample - "A non-parametric statistical test is based on a model that specifies only

very general conditions . . ." (Siegel & Castellan, 1988, p. 34). Additionally, the

distribution of scores required a "distribution-free approach" (Howell, 1997, p. 645).

Results of these tests and associated descriptive statistics (resulting from

exploratory data analysis) are reported in this chapter. In those analyses where a precise

probability was available it is quoted as a p value. An alpha level of < .05 significance

(denoted by a single asterisk - *) was used for all statistical tests. Significance levels

are reported at three decimal places.

171

Chapter 6: Results

6.4 HYPOTHESIS AND ASSOCIATED QUESTIONS

As mentioned earlier, the major hypothesis is that toddlers engaged in free play

with older, preschool-aged peers (either in mixed-age play or dyad play) will exhibit

symbolic-play skills in each of the four dimensions of symbolic play identified by

Westby (1991) - 'decontexutalisation', 'thematic content', 'organisation of themes' and

'self-other relationships'- more frequently and at a higher level, than when they are

engaged in free play with their same-age peers.

In conjunction with the major hypothesis, the following questions were also

considered:



symbolic play?


of the four dimensions of symbolic play identified by Westby (1991) —

'decontexutalisation', 'thematic content', 'organisation of themes' and

'self-other relationships'- more frequently and at a higher level than


c) Does the long-day child-care environment, which includes the quality

of care provided - the program of activities implemented, the materials

and equipment provided, staff training and qualifications, and other

available resources - support children's engagement in symbolic play?

6.5 A PRELIMINARY EXPLORATION OF THE DATA

The statistics package 'MLNtTAB' (Student Edition of Minitab for Windows

and Minitab 12) was used to compute the following descriptive statistics. Analyses of

these data are reported later in this chapter.

6.5.1 Descriptive Statistics: Means, Standard Deviations and Medians for the

Frequency of Symbolic-Play Events

The means, standard deviations and medians for the frequencies of toddler

participants' symbolic-play events in each condition .are presented in Table 6.3. In the

dyad play sessions four of the toddler participants stayed for less than 20 minutes out of

the possible 30 minutes: this could have resulted in fewer symbolic-play events being

172

Chapter 6: Results

observed. Thus a weighted average frequency of events per 30 minutes has also been

calculated for the dyad-play condition.

Table 6.3

Frequencies of Symbolic-Play Events in Three Conditions

Condition

Same-age

Mixed-age

Dyad

*Dyad

*Frequencies for symbolic-play events for dyad condition were weighted for length of observation period for those truncated by participants.

6.5.2 Descriptive Statistics: Means, Standard Deviations and Medians for the

Highest Levels of Symbolic Play

The means and standard deviations for the participants' highest levels in each

dimension of symbolic play - 'decontextualisation', 'thematic content', 'organisation of

themes' and 'self-other relationships' - in each of the three conditions, same-age play,

mixed-age play and dyad play, are presented in Table 6.4.

N

24

24

24

24

Mean

7.37

7.08

20.88

0.85

Standard Deviation

6.80

7.49

9.57

0.35

Medial

6.50

4.50

21.00

0.80

173

Chapter 6: Results

Table 6.4

The Highest Levels in Each Dimension of Symbolic Play in Each Condition (N = 24)

Dimension

'Decontextualisation'

Condition

Same-age

Mixed-age

Dyad

Mean

1.63

1.38

1.83

Standard Deviation

#2.04

1.84

1.90

'Thematic content' Same-age

Mixed-age

Dyad

1.83

1.67

2.00

1.40

1.37

0.00

'Organisation of themes'

'Self-other relationships'

Same-age

Dyad

1.63

2.58

1.28

Mixed-age

Dyad

Same-age

Mixed-age

1.58

2.17

1.80

1.75

1.38

0.48

1.18

1.45

1.06

#Standard Deviation larger than Mean as skewed distribution. (Dimension One 'decontextualisation' - has only 2 classes [1 or 6].)

6.5.3 Descriptive Statistics: Means and Standard Deviations for the Most-

Frequently Occurring Levels of Symbolic Play

The means and standard deviations for the most-frequently occurring levels of

each dimension of symbolic play ('decontextualisation', 'thematic content',

'organisation of themes' and 'self-other relationships'), exhibited by participants in each

condition (same-age play, mixed-age play and dyad play), are listed in Table 6.5.

174

Chapter 6: Results

Table 6.5

The Most-Frequently Occurring Levels in Each Dimension of Symbolic Play in

Each Condition (N = 24)

Dimension


'Thematic content'



Condition

Same-age

Mixed-age

Dyad

Same-age

Mixed-age

Dyad

Same-age

Mixed-age

Dyad

Same-age

Mixed-age

Dyad

Mean

1.21

1.17

1.00

1.33

1.38

1.46

1.33

1.33

1.58

1.50

1.50

1.88

Standard Deviation

*1.53

1.55

0.00

0.82

1.01

0.51

0.92

1.01

0.50

0.83

1.02

0.34

*Standard Deviation larger than Mean as skewed distribution. (Dimension One 'decontextualisation' - has only 2 classes [1 or 6].)

6.6 THE FREQUENCY OF SYMBOLIC-PLAY EVENTS

The first component of the major hypothesis was that toddler participants would,

when engaged in free play with older peers (in either mixed-age play or dyad play),

exhibit symbolic-play behaviour more frequently than toddlers w h o were not engaged in

free play with older peers.

The frequency of symbolic-play events that each participant engaged in during

the three different conditions under which they were observed was tallied. Friedman's

175

Chapter 6: Results

Rank test was then used to compare the frequency of symbolic-play acts (or events) in

each of the three conditions (same-age play, mixed-age play and dyad play). These data

are explored and discussed below.

The hypothesis, that toddlers would exhibit symbolic-play behaviours more

frequently in the company of older peers, was partially supported. A highly significant

difference was found in Condition Three (dyad play) - S = 17.80, df = 2, p = .000. ('S'

denotes the Friedman's Rank Test statistic; 'df denotes the degrees of freedom.)

Thus participants engaged in more acts of symbolic play in Condition Three

(dyad play) than in either Condition One (same-age play) or Condition T w o (mixed-age

play). These results are presented in Table 6.6.

Table 6.6

The Frequency of Symbolic-Play Events in Three Conditions

Condition

Same-age

Mixed-age

Dyad

N

24

24

24

Median

7.833

7.167

20.500

S u m of Ranks

39

40.5

*64.5

^denotes significance at p < .05.

Wilcoxon's Matched-Pairs Signed-Ranks test was used to show the

magnitude and direction of the results of Friedman's Rank test on the related samples.

The test showed that participants exhibited significantly more events of symbolic play

in Condition Three than in either Condition One or Condition Two. Results were as

follows:

- Condition Three (Dyad) > Condition One (Same-age): p = .000.

- Condition Three (Dyad) > Condition T w o (Mixed-age): p = .000.

6.6.1 The Relationship Between Birth Order and the Frequency of Symbolic-Play

Events

One of the questions posed, as an adjunct to the major hypothesis, was whether

first-born toddlers, when engaged in free play with older peers (whether mixed-age play

or dyad play), would exhibit symbolic-play activity more frequently than toddlers with

older siblings. The Kruskal-Wallis one-way analysis of variance by ranks was used to

176

Chapter 6: Results

examine the relationship between participants' birth order and the frequency of

symbolic-play events.

Significant results were not found. However, in the mixed-age play condition a

p value of .061 was recorded for first-born participants - this approaches significance.

Results are presented in Table 6.7.

Table 6.7

The Relationship Between Birth Order and the Frequency of Symbolic-Play

Events in Three Conditions

N o sibling Sibling Kruskal Wallis

N 14 10 Statistic

(H)

Condition Sum of ranks Sum of ranks

df p

Same-age 12.2 12.9 (0.07) 0.792

1

Mixed-age 14.8 9.3 (3.58) 0.061

1

Dyad 13.5 11.1 (0.67) 0.412

1

6.6.2 The Relationship Between Age Group and the Frequency of Symbolic-Play

Events

Participants were grouped, according to age as follows: 17-19 months; 20-22

months; 23-26 months; and 27-31 months. These groupings were largely in accordance

with those identified by Westby (1991) as likely to reflect a change in symbolic-play

activity.

The relationship between the participants' age group and the frequency of

symbolic play in the three conditions (same-age play, mixed-age play and dyad play)

was found using the Kruskal-Wallis test. This showed that children aged 20 to 22

months exhibited less symbolic-play activity in Conditions One and Three (same-age

177

Chapter 6: Results

play and dyad play) than in Condition T w o (mixed-age play) than the participants in the

other three age groups. Findings are summarised in Table 6.8.

Table 6.8

The Relationship Between Age Group and the Frequency of Symbolic-Play Events

in each Dimension in Three Conditions

Age group

(months)

N

Condition

17-19 20-22 23-26 27-31

Kruskal-

Wallis

statistic

(H)

Ave.rank Ave.rank Ave. rank Ave. rank df

Same-age

Mixed-age

Dyad

16.2

15.3

7.3

5.5

9.5

8.8

15.5

12.3

16.8

13.3

12.3

15.7

8.51

3

2.0

3

7.96

3

*0.037

0.573

H0.047

*denotes significance at p < .05

6.7 THE HIGHEST LEVELS OF SYMBOLIC PLAY

The Friedman's Rank test was then employed to compare the highest level of

each dimension ('decontextualisation', 'thematic content', 'organisation of themes' and

'self-other relationships') of symbolic play observed in each condition (same-age play,

mixed-age play and dyad play).

The hypothesis that toddlers would exhibit higher levels of symbolic-play

behaviours when engaged in free play with an older peer was supported. In dyad play

(Condition Three) Friedman's Rank test indicated that the differences between the

means for two dimensions of symbolic-play behaviour were significant. Friedman's

Rank test showed that there were significant differences amongst the conditions for

Dimension Three ('organisation of themes') - S = 8.52, df = 2, p = .014, and for

178

Chapter 6: Results

Dimension Four ('self-other relationships') - S = 7.73, df = 2, p = .021. These results

are presented in Table 6.9.

In summary, participants exhibited higher scores in dyad play than in same-age

play or mixed-age play for both 'organisation of themes' (Dimension Three) and 'self-

other relationships' (Dimension Four).

There were no significant differences amongst the conditions for

'decontextualisation' (Dimension One) or 'thematic content' (Dimension Two).

Table 6.9

The Highest Levels of all Dimensions of Symbolic Play in Three Conditions (N= 24)

Dimension Condition Estimated Median

1.00

1.00

1.00

2.00

2.00

2.00

1.33

1.17

2.00

2.00

2.00

2.00

Sum of Ranks

46.5

43.5

54.00

47.5

43.5

53.0

44.5

42.0

*57.00

45.5

42.5

*56.0


'Thematic content'

Same-age

Mixed-age

Dyad

Same-age

Mixed-age

Dyad

'Organisation of themes' Same-age

Mixed-age

Dyad

'Self-other relationships' Same-age

Mixed-age

Dyad


179

Chapter 6: Results

Once again Wilcoxon's Matched-Pairs Signed-Ranks test was employed to test

for the degree and direction of the results of Friedman's Rank test for the highest levels

of participants' symbolic play on related samples. The test showed that participants

exhibited significantly higher levels of Dimension Three ('organisation of themes')

symbolic-play behaviour in Condition Three (dyad play) than in either Condition One

(same-age play) or Condition T w o (mixed-age play).

Wilcoxon's test also showed that participants exhibited significantly higher

levels of Dimension Four ('self-other relationships') symbolic-play behaviour in

Condition Three (dyad play) than in either Condition One (same-age play) or Condition

Two (mixed-age play). These results are displayed below:

- Highest Dimension Three ('organisation of themes'):

. Condition Three (Dyad play) > Condition One (Same-age play)

- p = .023;

. Condition Three (Dyad play) > Condition T w o (Mixed-age play)

- p = .027.

- Highest Dimension Four ('self-other relationships'):

. Condition Three (Dyad play) > Condition One (Same-age play)

- p = .038;

. Condition Three (Dyad play) > Condition T w o (Mixed-age play)

- p = .015.

6.7.1 The Relationship Between Birth Order and the Highest Level of Symbolic

Play

The Kruskal-Wallis test was used to examine the relationship between

participants' birth order and the highest level of symbolic play. The prediction that

first-born toddlers would exhibit higher levels of symbolic play than later-born toddlers

was supported. In mixed-age play (Condition Two), higher levels of symbolic play in

all four dimensions: 'decontextualisation' (H = 3.77, df = 1, p < .05) 'thematic content'

(H = 4.06, df =1, p <.05); 'organisation of themes' (H = 4.77, df = 1, p < .05); and 'self-

other relationships' (H = 4.39, df = 1, p < .05) were evident for participants who were

first-borns. In dyad play (Condition Three) a p value of .070 for Dimension One

('decontexutalisation') for first-born participants approaches significance.

180

Chapter 6: Results

Thus first-borns scored significantly higher than later-borns in each of the four

dimensions of symbolic play in Condition T w o (mixed-age play). N o other significant

results were indicated for the relationship between birth order and highest levels of

symbolic play. These results are presented in Table 6.10.

Table 6.10

The Relationship Between Birth Order and the Highest Level of Symbolic Play in

each Dimension in Three Conditions

Dimension

N

No sibling

14

Sibling

10

Condition Sum of ranks S u m of ranks

Kruskal-

Wallis

statistic

(H)

df

'Decontextualisation' Same-age

Mixed-age

Dyad

'Thematic content'

'Organisation of

themes'

Same-age

Mixed-age

Dyad

Same-age

Mixed-age

Dyad

'Self-other relationships' Same-age

Mixed-age

Dyad

12.9

14.5

13.9

12.6

14.7

12.5

12.5

15.0

13.6

12.4

14.8

13.9

11.9

9.6

10.5

12.4

9.4

12.5

12.4

9.1

11.0

12.7

9.2

10.6

(0.14)

1

(3.77)

1

(3.29)

1

(0.01)

1

(4.06)

1

(0.00)

1

(0.00)

1

(4.77)

1

(2.34)

1

(0.02)

1

(4.39)

1

(2.16)

1

0.709

*0.053

0.070

0.921

*0.044

1.000

0.976

*0.029

0.127

0.883

*0.036

0.142

* denotes significance at p < .05

** denotes significance if taken to the conventional two decimal places of p < .05.

181

Chapter 6: Results

6.7.2 The Relationship Between Age Group and the Highest Level of Symbolic Play

The Kruskal-Wallis test revealed significant differences existed between the

means for participants in the second age group (20 to 22 months) when compared to

other age groups. These participants showed significantly lower levels of symbolic play

than other age groups in three dimensions in Condition One (same-age play) as follows:

Dimension One ('decontextualisation') - H = 8.94, df = 3, p < .05; Dimension T w o

('thematic content') - H = 10.26, df = 3, p < .05; and Dimension Four ('self-other

relationships') - H = 8.73, df = 3, p < .05. In Dimension Three, ('organisation of

themes') the value of p (.062) is approaching significance. These results are shown in

Table 6.11. Significant differences were not found in the mixed-age play condition or

the dyad play condition.

182

Chapter 6: Results

Table 6.11

The Relationship Between Age Group and the Highest Level of Symbolic Play in

each Dimension in Three Conditions

Dimension

Age group

(months)

N

Condition

17-

19

20-

22

23-

26

27-

31

Ave.

rank

Ave.

rank

Ave.

rank

Ave.

rank

Kruskal-

Wallis

statistic

(H)

df

'Decontextualisation' Same-age 13.0 6.3 16.2 14.3

Mixed-age 11.9 8.8 13.8 15.1

Dyad 10.5 10.5 14.5 13.9

(8.94)

3

(3.84)

3

(3-94)

3

*0.030

0.280

0.268

'Thematic content' Same-age 14.5 5.5 16.1 14.0

Mixed-age 11.5 9.5 13.5 14.9

Dyad 12.5 12.5 12.5 12.5

(10.26)

3

(2.70)

3

(0.00)

3

*0.016

0.440

1.000

'Organisation of themes' Same-age 14.5 6.1 15.6 13.9

Mixed-age 12.3 8.7 14.3 14.4

Dyad 11.0 11.0 15.1 12.6

(7.33)

3

(3.10)

3

(3.94)

3

0.062

0.377

0.269

'Self-other relationships' Same-age 14.0 6.7 14.0 15.1

Mixed-age 11.2 9.3 13.1 15.7

Dyad 11.7 15.5 11.8 11.1

(8.73)

3

(3.54)

3

(2.58)

3

*0.033

0.315

0.460


183

Chapter 6: Results

6.8 THE MOST-FREQUENTLY OCCURRING LEVELS OF SYMBOLIC PLAY

Finally, the Friedman's Rank test was used to examine the most-frequently

occurring levels for each dimension of symbolic play exhibited by each participant.

This information was deemed useful as a possible indicator for participants' potential

level of symbolic-play. N o significant differences were found between the groups for

any of the conditions.

6.8.1 The Relationship Between Birth Order and the Most-Frequently Occurring

Levels of Symbolic Play

The Kruskal-Wallis test was also used to examine the relationship between

participants' birth order and the most-frequently occurring levels of symbolic play. In

Condition T w o (mixed-age play), statistical significance was registered for the most-

frequently occurring levels of play shown by first-born toddler participants.

Significance was recorded in three dimensions of symbolic play: 'thematic content' (H

= 3.74, df = 1, p < .05),'organisation of themes' (H = 5.37, df = 1, p < .05) and 'self-

other relationships' (H = 6.00, df = 1, p < .05) as shown in Table 6.10. In addition in

Condition T w o (mixed-age play) the p value for Dimension One ('decontextualisation')

of .067 approaches significance. N o significant trends were recorded for birth order on

the most-frequently occurring levels of symbolic play in either of Conditions One and

Three (same-age play and dyad play). Thus first-born participants scored significantly

higher than later-born participants in three dimensions of symbolic-play behaviour in

Condition T w o (mixed-age play). These results are presented in Table 6.12.

184

Chapter 6: Results

Table 6.12

The Relationship Between Birth Order and the Most-Frequently Occurring Levels

of Symbolic Play in each Dimension in Three Conditions

No sibling Sibling Kruskal-

Wallis

N 14 10 statistic

(H)

Dimension Condition Sum of ranks S u m of ranks df p

'Decontextualisation' Same-age 13.0 11.8 (0.26) 0.607

1

Mixed-age 14.4 9.9 (3.37) 0.067

1

Dyad 12.5 12.5 (0.00) 1.000

1

'Thematic content' Same-age 12.8 12.1 (0.05) 0.821

1

Mixed-age 14.7 9.4 (3.74) **0053

1

Dyad 11.3 14.2 (1.33) 0.249

1

'Organisation of themes' Same-age 12.1 13.0 (0.10) 0.758

1

Mixed-age 15.1 8.8 (5.37) *0.021

1

Dyad 11.5 13.9 (0.92) 0.338

1

'Self-other relationships' Same-age 12.5 12.4 (0.00) 0.971

1

Mixed-age 15.1 8.9 (6.00) *0.014

1

Dyad 12.3 12.8 (0.09) 0.759

1

* denotes signficance at p < .05

** denotes significance if taken to the conventional two decimal places of p < .05.

185

Chapter 6: Results

6.8.2 The Relationship Between Age Group and the Most-Frequently Occurring


The Kruskal-Wallis one-way analysis of variance by ranks was used to

examine the relationship between age group and the most-frequently occurring levels of

symbolic play. Results are presented in Table 6.13. The significant results were due to

the lack of activity (low ranks) in the second age group - that is children aged 20 to 22

months - in same-age play (Condition One). Significant results were shown in

Dimension One - 'decontextualisation' - (H = 8.93, df = 3, p < .05) and Dimension

Two -'thematic content' - (H =7.80, df = 3, p < .05). In addition p values approaching

significance were recorded for Dimension Three (p = .074), and Dimension Four (p =

.056).

Thus participants in the second age group scored significantly lower than other

age groups in two dimensions of symbolic-play behaviour in Condition One (same-age

play).

186

Chapter 6: Results

Table 6.13

The Relationship Between Age Group and the Most-Frequently Occurring Levels

of Symbolic Play in each Dimension in Three Conditions

Dimension


'Thematic content'



Age group

(months)

N

Condition

Same-age

Mixed-age

Dyad

Same-age

Mixed-age

Dyad

Same-age

Mixed-age

Dyad

Same-age

Mixed-age

Dyad

17-

19

5

Ave.

rank

14.0

12.3

12.5

14.2

9.6

11.8

16.3

11.7

12.7

16.0

11.8

11.6

20-

22

6

Ave.

rank

6.7

9.0

12.5

6.4

10.5

9.0

6.6

9.3

9.5

7.3

9.8

14.0

23-

26

6

Ave.

rank

15.6

14.2

12.5

13.3

13.8

11.0

14.7

14.2

11.5

12.8

13.8

10.0

27-

31

7

Ave.

rank

13.8

14.2

12.5

15.9

15.2

17.3

13.0

14.4

15.8

14.1

14.2

14.0

Kruskal-

Wallis

statistic

(H)

df

(8.93)

3

(3.22)

3

(0.00)

3

(7.80)

3

(2.98)

3

(6.70)

3

(6.92)

3

(2.38)

3

(3.72)

3

(7.58)

3

(2.13)

3

(4.31)

3

P

*0.030

0.359

1.000

*0.050

0.394

0.082

0.074

0.497

0.294

0.056

0.545

0.230


187

Chapter 6: Results

6.9 THE COMPLEXITY OF SYMBOLIC PLAY

The statistical analyses conducted thus far responded to questions based on

whether or not participants exhibited higher levels of symbolic play, and more

frequently, when engaged in free play with a preschool-aged peer - either in mixed-age

play or in dyad play. Analyses for the most-frequently occurring level of symbolic play

(the level of symbolic play in which participants most often engaged) was also

undertaken. Additionally, relationships between the participants' birth order and age

group and the frequency of their symbolic play, their highest levels of symbolic play

and their most-frequently occurring levels of symbolic play were examined.

The theoretical stance offered by Westby (1991) supported the use of Logistic

Regression Analysis. As mentioned earlier in Chapter One (1.8) Westby (1991)

identified that an even pattern was observable in the development of young children's

symbolic play resulting in similar levels across the four dimensions (or behaviours) -


relationships' - of such play. Nevertheless, Westby (1991) recognised that differences

in dimension levels would be discernible as each reflects the varied social, language and

cognitive aptitudes of the players. A further consideration is that the activity and

content of the four symbolic-play dimensions are also dependent upon the experience

that children bring with them to the play situation. (See Appendix Gii.)

In order to undertake a Logistic Regression Analysis it was deemed necessary to

categorise each participant's symbolic-play activity as either above, or equal to or

below, the expected age level. (It may be remembered that this binary variable was

named 'complexity' earlier in Chapter Five, 5.9.6.)

6.9.1 Complexity and Participants' Ages

Logistic Regression Analysis allowed the investigation of the highest level at

which a participant engaged in symbolic play and whether this was above the level

predicted by Westby's (1991) Scale for Assessing Children's Pretend Play for a child of

a stated age. 'Complexity' as stated earlier was denoted as either: a) greater than the

level of symbolic play expected for the participant's age according to Westby's (1991)

scale; or b) less than or equal to the expected level of symbolic play for participants of

that age. Age was the independent variable.

Logistic Regression Analysis showed that participants in the youngest age group

(17 to 19 months) exceeded the expected level of complexity in symbolic play in

188

Chapter 6: Results

Dimension T w o ('thematic content') - p < .05 and Dimension Four ('self-other

relationships') - p < .05 in Condition One (same-age play). The expected levels for

Dimension One ('decontextualisation') and Dimension Three ('organisation of themes')

- were not exceeded in Condition One (same-age play).

In Condition T w o (mixed-age play) the expected level for Dimension One

('decontextualisation') was exceeded - p < .05.

In Condition Three (dyad play) the expected level was exceeded for: Dimension

One ('decontextualisation') - p < .05; Dimension Three ('organisation of themes') - p <

.05; and Dimension Four ('self-other relationships') - p < .05. (Age in months was the

continuous variable.) These results are detailed in Table 6.14.

Table 6.14

Significance of Age in Logistic Regression of Complexity on Age for Participants

Aged 17-19 Months Exceeding Expected Level of Each Dimension of Symbolic

Play (according to Westby, 1991), in Each Condition

Condition 1. Same-age 2. Mixed-age 3. Dyad play play play

Dimension 1.Decontextualisation 0.058

2. Thematic content *0.003

3. Organisation of 0.062 themes

4.Self-other *0.002 relationships

^denotes significance at p < .05 ** denotes no variability between children all '0' ('less than or equal to' level expected)

In summary, use of Logistic Regression Analyses showed that the chance of

exceeding the 'expected' level of symbolic-play activity in Dimension T w o ('thematic

content') and Dimension Four ('self-other relationships') - according to Westby (1991)

- in same-age play (Condition One) depended on the age of the child (p < .05). This

relationship appears most strongly in the youngest age group (17 to 19 months). In

same-age play (Condition One) the chance of exceeding the 'expected' level was less

dependent on age for Dimension One ('decontextualisation') - p = .058 and Dimension

*0.049

0.215

0.215

0.355

*0.027

**

*0.000

*0.000

189

Chapter 6: Results

Three ('organisation of themes') - p = .062. For Dimension One ('decontextualisation')

it is the older age groups, and for Dimension Three ('organisation of themes') it is the

youngest age group, that are most likely to exceed their 'expected' level.

In mixed-age play (Condition T w o ) and dyad play (Condition Three) there are

similar patterns. In mixed-age play the age relationship is statistically significant only

for Dimension One ('decontextualisation') - p < .05. In dyad play the results are highly

significant: Dimension One ('decontextualisation') - p < .05; Dimension Three

('organisation of themes') - p < .05; and Dimension Four ('self-other relationships') - p

<.05. (See Table 6.14.)

The complexity (again in terms of either: greater than that expected; or less than

or equal to that expected) of participants' symbolic-play activity in each of three

dimensions - 'thematic content', 'organisation of themes' and 'self-other relationships'

- was explored further using Logistic Regression Analysis. (As Dimension One

['decontextualisation'] has only two levels - 1 or 6 - it was deemed as not useful to

explore using this method.)

6.9.2 Explanatory Variables for Complexity

To account for enhanced complexity of symbolic play in the three dimensions of

symbolic play ('thematic content', 'organisation of themes' and 'self-other

relationships'), the following variables were considered as possible explanatory

variables:

- same-age play versus not same-age play (that is, mixed-age play and

dyad play);

- dyad play versus not dyad play (that is same-age play and mixed-age

play)

- the materials for play provided by the L D C C C ;

- the rating awarded the L D C C C on the Environment Rating Scale;

- the level of education achieved by the participants' parents;

- the participants' weekly attendance at the L D C C C .

Logistic Regression Analysis was then used to calculate the 'likelihood' (or odds

ratio) of the complexity of a participant's symbolic play being greater than that expected

for their age according to Westby's (1991) Scale for Assessing Children's Pretend Play.

190

Chapter 6: Results

Use of the 'odds ratio' statistic "makes clear the degree to which one variable influences

another" (Howell, 1997, p. 159).

For complexity in Dimension T w o ('thematic content') the useful variables were

dyad play (Condition Three) and provision of play materials. Results showed that

participants from L D C C C s where the provision of play materials is unsatisfactory are

5.5 times (p = .034) as likely to show complex play than children from L D C C C s where

the provision of play materials is satisfactory. Participants are 2.2 (p = .048) times as

likely to show complex play in dyad play (Condition Three) than in same-age play

(Condition One) or mixed-age play (Condition Two). These two variables, provision of

play materials and dyad play, are statistically independent - that is they do not interact.

For complexity in Dimension Three ('organisation of themes') the explanatory

variables were provision of play materials and the days participants attended L D C C C -

whether three days or four or five days. Again, interestingly, participants from

L D C C C s where the provisions of play materials are unsatisfactory are 12.5 times (p =

.004) as likely to engage in complex play than children in L D C C C s where the provision

of play materials is satisfactory. Participants attending L D C C C for four or five days per

week are 14 times (p = .066) as likely to engage in complex play as those attending for

three days per week. These two influences - provision of play materials and days

attended - do not interact: they are independent. (These findings are discussed in detail

in the next chapter.)

The predicting variables to explain complexity in Dimension Four ('self-other

relationships') are dyad versus not-dyad and unsatisfactory provision of play materials

versus satisfactory provision of play materials. Again, the results were in contradiction

to the predictions and are discussed in Chapter Seven. Results of analysis showed that

participants from L D C C C s where the provision of play materials is unsatisfactory are

5.3 times (p = .038) as likely to show complex play than children from L D C C C s where

the provision of play materials is satisfactory. Participants are 2.2 times (p = .047) as

likely to show complex play in dyad play (Condition Three) than in same-age play

(Condition One) or mixed-age play (Condition Two). Again the two explanatory

variables are independent and do not interact.

These results, explaining complexity in three dimensions are presented in Table

6.15.

191

Chapter 6: Results

Table 6.15

Explanatory Variables for the Likelihood (or Odds Ratio) for Complexity to Occur

in Three Dimensions of Symbolic Play ('thematic content', 'organisation of

themes' and 'self-other relationships')

Dimension Variable Likelihood (odds

ratio)

p value

2. 'Thematic content'

3. 'Organisation of

themes'

Unsatisfactory provision of play

materials

Condition 3 (Dyad play)


materials

4 or 5 days weekly attendance

5.5

(p = .034)

2.2

(p = .048)

12.5

(p = .004)

14

(p = .066)

4. 'Self-other

relationships'


materials

Condition 3 (Dyad play)

5.3

(p = .038)

2.2

(p = .047)

192

Chapter 6: Results

6.10 THE RELATIONSHIP BETWEEN GENDER AND SYMBOLIC PLAY

While the relationship between gender and the frequency, level and complexity

of symbolic play was not identified initially as an independent variable of this study, it

was nevertheless considered useful to investigate. N o significant gender differences

were revealed in any results.

6.11 SUMMARY OF STATISTICAL RESULTS

Key to the three conditions under which children were observed and video-taped

and the four dimensions - or behaviours - of symbolic activity:

- Condition One - Same-age play;

- Condition T w o - Mixed-age play;

- Condition Three - Dyad play;

- Dimension One - 'Decontextualisation';

- Dimension T w o - 'Thematic content';

- Dimension Three - 'Organisation of themes';

- Dimension Four - 'Self-other relationships'.

6.11.1 Results using Friedman's Rank Test

The Frequency of Symbolic-Play Events:

- Significant differences between the means for the frequency of

symbolic-play events were found in Condition Three (p = .000).

The Highest Levels of Symbolic Play:

- Significant differences between the means for the participants'

highest levels of Dimension Three (p = .014), and Dimension Four

(p = .021) were found in Condition Three.

The Most-Frequently Occurring Levels of Symbolic Play:

- N o significant differences between the means for the participants'

most-frequently exhibited levels in any of the four dimensions of

symbolic-play behaviour were found between: Condition One and

Condition Two; Condition One and Condition Three; and Condition

T w o and Condition Three.

193

Chapter 6: Results

6.11.2 Results using the Kruskal-Wallis Test

The Relationship Between Participants' Birth Order and the Frequency of

Symbolic-Play Events:

- N o significant differences between the means for the frequency of

symbolic-play events were found in any of the four dimensions in any

condition. However, in Condition Two, for first-born participants a

p value of .061 in Dimension One approaches significance.

The Relationship Between Participants' Birth Order and the Highest Levels of

Symbolic Play:

- Significant differences between the means of the highest levels of

each of the four symbolic-play behaviours were found for first-born

participants in Condition T w o only: Dimension One (p = .053);

Dimension T w o (p = .044); Dimension Three (p = .029); and

Dimension Four (p = .036).

The Relationship Between Participants' Birth Order and the Most-Frequently

Occurring Levels of Symbolic Play:

- Significant differences between the means of the most-frequently

occurring levels of symbolic-play behaviours in three dimensions were

found for first-born participants in Condition T w o only: Dimension

T w o (p =.053); Dimension Three (p = .021); and Dimension Four

(p = .014).

The Relationship Between Participants' Age Group and the Frequency of

Symbolic-Play Events:

Though unanticipated earlier in the project, the following two results have

proved useful for discussion in relation to the major hypothesis and additional

questions:

- Significant differences between the means of the frequency of

symbolic-play events in all four dimensions were found for

participants aged 20 to 22 months in Condition One (p = .037) and

Condition Three (p = .047).

194

Chapter 6: Results

The Relationship Between Participants' Age Group and the Highest Levels of

Symbolic-Play:


each of the four symbolic-play behaviours were found for participants

aged 20 to 22 months in Condition One only: Dimension One

(p = .030); Dimension T w o (p = .016); and Dimension Four (p = .033).

A p value of .062 in Dimension Three approaches significance.

The Relationship Between Participants' Age Group and the Most-Frequently

Occurring Levels of Symbolic-Play:

- Significant differences between the means of the most-frequently-

occurring levels of symbolic play were found for participants aged 20 to

22 months in Condition One: Dimension One (p = .030) and,

Dimension T w o (p = .050). A p value of .074 for Dimension Three and

a p value of .056 for Dimension Four approach significance. (There

were no significant differences for the other age groups.)

6.11.3 Results using Logistic Regression Analysis

The Relationship Between Age and the Complexity of Symbolic Play

A relationship between age and the complexity of symbolic play (symbolic

activity that is either 'greater than' or 'less than or equal to' that expected) was found in:

- Condition One in Dimensions T w o (p = .003) and Four (p = .002)

and the p value of .062 for Dimension One approaches significance;

- Condition T w o in Dimension One (p = .049);

- Condition Three in three dimensions - Dimension One (p = .030),

Dimension Three (p = .000) and Dimension Four (p = .000).

Variables Useful to Explain the Likelihood (the Odds Ration) of Participants'

Showing Complexity of Symbolic Play

- Variables identified as useful to explain the likelihood (the odds

ratio) of participants' showing complexity of symbolic play in

Dimension T w o were Condition Three - dyad play - (likelihood 2.2,

p = .048) and unsatisfactory provision of play materials (likelihood

5.5, p = .034).

195

Chapter 6: Results


ratio) of participants showing complexity of symbolic play in

Dimension Three are a weekly attendance at L D C C C of four or more

days (likelihood 14, p = .066) and unsatisfactory provision of play

materials (likelihood 12.5, p = .004).

-Variables identified as useful to explain the likelihood (the odds ratio)

of participants showing complexity of symbolic play in Dimension Four

are Condition Three -dyad play (likelihood 2.2, p = .038) and

unsatisfactory provision of play materials (likelihood 5.3, p = .047).

(Data relating to the variables referred to in the results arising from Logistic

Regression Analysis are discussed in the next section, 6.12.)

6.12 ANALYSIS OF DESCRIPTIVE DATA

It will be recalled that additional instruments including an Environment Rating

Scale, a Parent Questionnaire and a Staff Questionnaire were employed in this project.

Information was also collated on staff-child ratios and toys, materials and equipment

supplied by centre staff for both Conditions One and Two. These data supplemented

those gathered from the Environment Rating Scale. The Staff Questionnaire elicited

attitudes toward, and value placed by participating staff on, children's play. It also

allowed the identification of levels of training and category of qualifications in Early-

Childhood Education achieved by staff members.

Each of these components - the Environment Rating Scale; staff-child ratios; the

provision of toys, materials and equipment for play; and staff beliefs, attitudes and

qualifications - contributes to the overall quality of care-giving. In addition such

information assists the researcher to respond to the following question - posed in

addition to the major hypothesis, as mentioned in Chapter One, 1.4.

Does the long-day child-care environment, which includes the quality of care

provided - the program of activities implemented, the materials and equipment

provided, staff training and qualifications, and other available resources - support

196

Chapter 6: Results

children's engagement in symbolic play? These questions will now be addressed in

6.12.1.

6.12.1 Results of the Environment Rating Scale

As described in Chapter Five (5.8.4.1), the Environment Rating Scale used in

this study comprised twelve items adapted from the ITERS (Infant/Toddler

Environment Rating Scale) constructed by Harms, Cryer and Clifford (1990). Rating

was accomplished by means of a Likert scale (as is used for rating ITERS) ranging from

'(1) inadequate' to '(7) excellent'. For each item, the assessment of standard of care

giving in terms of quality was determined in the following manner: a rating up to 3 on

the Likert scale indicated the provision of a minimal standard of care - below that

judged by Harms et al. (1990) as adequate quality care-giving conditions; a rating of 4

denoted an adequate provision of care; a rating of 5 or above denoted provision of a

good quality care-giving environment with the top rating of 7 indicating provision of

excellent or high-quality care.

The maximum score attainable, using the modified Environment Rating Scale

described here, was 84. A total rating of between 46 and 50, indicating that the average

rating across the twelve items was approximately 4 (4 being above '3 - minimal1 but not

quite reaching '5 - good'), was nominated as the range for describing the centre's quality

of care-giving as adequate. Therefore a score of 45 or below resulted in the quality of

care provided being recognised as unsatisfactory. A score of 51 or more allowed a

centre's provision of care to be of 'good quality' and acknowledged as satisfactory. The

Environment Rating Scale completed independently by the researcher was used for the

above process.

Although data were collected from the 12 L D C C C s that took part in the research

project, complete sets of data were forthcoming from only eight centres. The

information in Table 6.16 relates to those eight centres only.

197

Chapter 6: Results

Table 6.16

Environment Rating Scale Scores and Accreditation

Items on rating scale

Centre Accreditation 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 Total U Id (in years) # score A

G

3

1

3

*N

3

**p

3

2

6

5

5

5

5

6

4

5

6

5

4

4

5

5

4

3.5

5

4

4

4

6

4.5

5

5

5

3

3

4

5

5

3

3

5

4

5

5

5

7

3.5

3

4

3

3

4

6

6

2

3.5

4

3

5

3

6

5.5

3

3

4

4

4

4

5

6

3

3.5

4.5

3

3.5

4.5

5

4.5

2.5

3

5

4

4

4

5

5.5

3.5

4

5

3

3.5

4

4.5

5

2

4

4

5

5

5

5

6

4

4

57.7

46

49

50.5

62.5

67

39.5

44.5

S

A

A

S

S

S

US

US

• ^ • ^ — — — • — • m m m m m ^ m m a m l ^ m m l m m m m l l m m l t a ^ l m ^ l l m l m ^ ^ ,111. ( M I I > M I < < t M I W M I I I I I , I I I M I I t l l l < < I I I B M ^ I M I W ^ ^

U denotes unsatisfactory quality of care-giving A denotes adequate quality of care-giving G denotes good or better than good quality of care-giving # denotes only regulatable item *N denotes new centre yet to be assessed for accreditation **F denotes centre participated in accreditation process but failed to gain

accreditation

In summary: two LDCCCs were categorised as providing low-quality

(unsatisfactory) care; two centres were classed as providing adequate care; and the

remaining four centres acknowledged as providing good-quality (satisfactory or better)

care. It is worthy of note that the one centre (Centre Nine), which failed to meet

accreditation standards and therefore failed to gain any period of accreditation, achieved

the highest rating on the Environment Rating Scale. The Environment Rating Scale

score of 67 achieved by Centre Nine indicates a provision of quality care bordering on

the 'very good' classification. This centre, known to the researcher before data

collection took place, had undergone extensive staff changes and six months after data

collection was finalised this centre achieved three years accreditation - the longest

period awarded.

198

Chapter 6: Results

In some cases, there were substantial differences between the ratings indicated

by the researcher's independent assessment and those awarded by the Director of the

L D C C C . The Director completed the Scale in the company of the researcher, who,

however, recorded the Director's own assessment and actively sought not to influence

her decisions. The differences, between the two scores awarded - ranging from 7 to 32

- were always in favour of the centre.

Interestingly, the smallest difference in scores occurred for Centre 9 - the

L D C C C identified by the researcher's completed Environment Rating Scale as offering

the highest-quality care of any participating centre. The greatest differences in the two

Environment Scale scores were those for L D C C C s rated by the researcher as providing

either unsatisfactory or adequate care only. In summary it appeared where the

Directors' ratings differed from the present author's the former were always high. The

centres offering poorer quality care were rated higher by the Directors - the more so the

as the quality was poorer.

Given the current emphasis on the Quality Improvement and Accreditation

System (QIAS), which involves child-care staff in a process of both self-evaluation and

preparation for an external review, it is perhaps likely that differences in perception

between the Director and researcher would occur. It seems plausible that Directors

would perhaps view their centres subjectively and consequently more favourably in

contrast to the researcher's more objective approach.

6.12.2 Additional Components of the Long-Day Child-Care Environment

Other data relating to both the children's play environment and the quality of that

environment were also collected. These data included the toys, materials and equipment

provided for children's use in Conditions One and T w o - same-age play and mixed-age

play. The number of children present in those two conditions was also recorded

together with the staff-child ratios.

6.12.2.1 Toys, Materials and Equipment

A n assessment of the toys, materials and equipment provided by the long-day

child-care staff and the suitability of those provisions in supporting the development of

children's symbolic play was conducted in the following manner. During the data-

collection period the equipment supplied by staff for the children's free-play outdoor

activities each day was noted. (As earlier stated a prerequisite for a centre to participate

199

Chapter 6: Results

in the study was that both the mixed-age free play and same-age free-play sessions took

place in the same outdoor environment with the same play materials, and that this

procedure was part of the centre's regular programming.)

The availability of toys, materials and equipment, was classified as one of the

following four categories: high realism, small representational, construction, and

realistic props for thematic play. These four categories are those suggested by Westby

(1991) as most appropriate for supporting the development of young children's

symbolic play. If materials belonging to one category were supplied on at least half the

occasions when data were collected then the provision rate of those materials was

considered 'satisfactory'. Failure to supply any category of toys, materials and

equipment for 49 percent or less of the observation periods resulted in a grading of

'unsatisfactory' provision (see Table 6.17). In some centres equipment belonging to one

category was not supplied during any of the observation periods. If two or more

categories of toys, materials and equipment were scored as being provided on at least

half of the occasions, the overall rating for provision of play materials awarded to that

centre was 'satisfactory'. It is important to remember that an 'unsatisfactory' grade for

any category indicates that toys, materials and equipment were provided, but irregularly

so. The number of different days the researcher visited centres to observe same-age

play and mixed-age play ranged from four to nine, the average being six and a half days.

In summary, as seen in Table 6.17, the toys, materials and equipment available

during the observed free-play sessions reveal that a high-realism (life-like) toy area was

provided in six centres on at least half the occasions when children were observed. In

one centre there was irregular provision of high-realism props, and in another centre the

researcher never, during the data collection period, observed the provision of these

materials.

200

Chapter 6: Results

Table 6.17

Percentage of Total Occasions of Provision of Toys, Materials and Equipment by

Centre

Centre Id

*High Realism

%

**Small representational

%

***Construction with small

representational

%

****

Realistic props for thematic play %

Overall rating

(Sat or Unsat.)

Rating on Environment

Scale US A G

11

12

50

0

33

63

67

63

83

67

0

75

50

13

0

38

0

0

0

75

67

75

67

13

0

56

83

0

17

0

67

0

0

0

Sat

Sat

Sat

Sat

Sat++

Unsat

Unsat

Sat

G

G

G

US

US

Unsat Sat US A G *

**

***

****

S++

denotes unsatisfactory provision of play materials denotes satisfactory provision of play materials denotes unsatisfactory provision of care-giving denotes adequate provision of care-giving denotes good or better provision of care-giving denotes area equipped with familiar high-realism (life-like) props. denotes area equipped with small representational toys. denotes area equipped with small representational toys in vicinity of construction materials denotes an area with realistic props to support thematic play. denotes centre with highest provision of toys, materials and equipment likely to support the development of pretend play.

6.12.2.2 Staff-Child Ratios in L D C C C s

Staff-child ratios were also collated for this project. This ratio is considered to

be a 'structural' component of the LDCCC and is an important adjunct to the process

dimensions when assessing the quality of care-giving in the long-day child-care

environment (Scarr et al., 1994). Information required to calculate staff-child ratios was

gathered at the start and completion of each period of video-taped observations in

201

Chapter 6: Results

Conditions One and T w o (same-age play and mixed-age play). The numbers of staff

and children in the play environment were counted and results recorded. It is

recognised that such a method results in an approximation only of staff-child ratios -

judged as adequate information for this project.

The current project was undertaken in Sydney, N e w South Wales in Australia,

where regulations exist for staff-child ratios according to the age of the child. Children

under two years of age are required to be in groups of five with one adult; one adult to

eight aged two to three and one adult to 10 children over three years old (New South

Wales State Government, 1996). Clearly when children are in mixed-age groups a

broader age group is represented, and overall a larger number of children is able to be

catered for by fewer staff and adherence to regulations still maintained.

During the data collection period ratios at times fell outside the N e w South

Wales regulations. This usually occurred under extenuating circumstances - for

example delayed staff arrivals or wide-spread sickness among staff. Generally though,

the average figures for staff-child ratios were in keeping with those required by

regulation.

6.12.3 An Analysis of Responses to Parent Questionnaire

The Parent Questionnaire was constructed to elicit information considered

essential to assist in the exploration and understanding of the results likely to arise from

the study. The questionnaire allowed an approximate assessment to be made of the

child's socio-economic background, the parents' knowledge and understanding of

development, and family leisure-time activities. It also contained the more pragmatic

questions including whether or not the project participant had siblings, and the main

language spoken at home. A n indication of why each child was placed in L D C C and

reasons for the choice of L D C C C were also requested.

The 91.5 percent response rate was high - 22 parents completed and returned the

document and two failed to do so.

6.12.3.1 Responses to Question 1 and 2

Responses to Question One (reasons for child's placement in L D C C ) were

categorised in the following four ways:

- indicating the prime reason for the child's placement in L D C C was to

allow the caregiver to return to work;

202

Chapter 6: Results

- indicating benefit to the child as the prime reason;

- citing both a return to work and benefits to the child;

- and finally any other reason.

Responses to Question T w o (reasons for choice of specific L D C C C ) were

categorised as:

- choice occurring through a 'philosophical' approach (for example a

recommendation by another parent or an elder sibling having attended

the centre);

- logistic reasons such as close proximity to the workplace or home;

- citing both a philosophical and logistic reason;

and any other reasons cited for choosing the L D C C C .

Responses to Questions One and Two are listed in Table 6.18.

Table 6.18

Frequencies and Percentages of Parents' Responses to Questionsl and 2

Question 1 Reason leading to placement of child in L D C C

Response Frequency

Percentage %

a) Adult's benefit (e.g. a return to

work)

19

86.5

b).Child's benefit (meet, play with

other children)

0

0

c) Both aandb cited

3

13.5

d) Other

0

0

Question 2 Reason for choice of L D C C C

Response Frequency

Percentage %

a)'Philosophical' (recommended or used satisfactorily

before)

8

36

b) Child's benefit

(meet, play with other children)

2

4.55

c) Both aandb cited

13

59

d) Other

4

0

203

Chapter 6: Results

While several responses to Question One, w h y placement was necessary,

included factors pertaining to both adults' and children's needs, no parent responded that

placement was purely for the benefit of the child. In all cases, the reason for use of the

L D C C C was the return to work of the 'child-caring' parent. In all but one case the

primary carer returning to work was female. While financial reasons for a return to the

workforce were by far the most c o m m o n statements, two responses citing resumption

and pursuit of career were made.

A number of responses included a reference to the lack of family or relatives,

either in the city environment or anywhere in Australia, available to care for the

children. It may be speculated from these responses that family, if available, would

have been the preferred choice of carer. However it is interesting to note that while a

return to the work-force was the prime reason for placement, few parents (13.5 percent)

made some comment in terms of favourable outcomes for their children. These

included opportunities to mix with peers, enjoy access to a wide range of activities and

play equipment and the belief that a 'centre-approach' would be enjoyed by the child.

One interesting comment suggested that the child's attendance at L D C C C not only

allowed the parent's return to work, but afforded the child some 'parent-free' time and

like-wise, the parent some 'child-free' time.

In response to Question T w o approximately 36 percent of parents cited the

referral or recommendation of the L D C C C as the first reason for their choice.

Proximity to home or workplace was the main consideration for only 4.55 percent of

parents. However overwhelmingly (59 percent) responses reflected a combination of

referral or recommendation by other parents or, as for five parents, a favourable

experience with an older child as being influential in the choosing of a L D C C C . Four

responses referred to placement availability in terms that suggested parents may have

experienced difficulty in finding vacancies for their children in L D C C C s . Quality was

cited in two responses but not elaborated upon.

Other factors were found to be important in the decision-making process, though

and these included fees which were mentioned as a positive consideration - presumably

in terms of being lower than those in other available centres - and also what parents

viewed as a 'well-designed' building. Placement availability and cleanliness were also

cited as favourable features.

Several parents referred to earlier unsatisfactory child-care experiences in terms

of provision for older children and that L D C C was now their preferred option. In this

204

Chapter 6: Results

group reference was made to the use of Family Day Care, house-keepers and nannies as

child-care providers. N o specific comments were made concerning the satisfaction or

otherwise with the Family Day Care service, but reference to 'in the home' carers such

as house-keepers and nannies focused on the unreliability of such arrangements.

Comments, in favour of L D C C such as the service offering a 'safety net' in terms of

care, lead the researcher to suggest that these early-childhood services are essentially

viewed as reliable - obviously a key factor for working parents.

N o specific reference to the presence of trained staff was made by any of the

respondents - although this notion could be contained in the statements naming quality

as a factor influencing choice. However, favourable comments included those about the

friendliness of staff, open channels of communication, or the pleasant 'happy'

atmosphere (described by one respondent as the 'feel' of the centre) parents' perceived

when visiting the centre.

6.12.3.2 Responses to Questions 3 and 4

Question Three required parents to briefly describe the development they would

expect to occur in their child in the ensuing six months. The following areas of child

development were listed for responses: cognitive or intellectual, physical, emotional,

social and other. Table 6.19 details these responses.

Question Four required each respondent to describe typical weekend leisure

activities that his or her child engaged in with parents, siblings, friends or others.

Responses to this question were rated as either positive or negative in terms of the play

'factor'. If language depicting play was used, the response was categorised as positive,

and conversely, if the notion of playful activity was in no way discernible in the

description, a negative rating was given. Table 6.19 details the responses to Questions

Three and Four.

205

Chapter 6: Results

Table 6.19

Frequencies and Percentages of Responses to Questions 3 and 4

Question 3

(Expectations for child's Appropriate developmental Appropriate developmental development over outcomes cited outcomes not cited subsequent six months)

Response Frequency 13 9

Percentage % 59 41

Question 4.

(Notion of play cited in Notion of play cited Notion of play not cited description of weekend leisure activities)

Response Frequency 16 6

Percentage % 73 27

Responses to these questions, rated as either positive or negative, were quite

difficult to categorise. However, it was deemed that for a response to be rated as

positive (a 'yes') some indication of the developmental nature of children's growth and

activity must be evident. A n example of this was one parent who stated that she

expected her child to gradually acquire the skills required for successfully playing with

other children as opposed to playing alone or in a solitary state. The wording of this

example tended to be the exception rather than the rule in terms of expressing the

notions of the progressive nature of developmental change. However, any indication of

an understanding of children's development was categorised as a positive response.

Responses not indicating an understanding of developmental change, for example, 'to

stop having temper tantrums', were categorised negatively. Responses which the

researcher perceived as borderline were rated favourably as positive.

In 59 percent of cases, positive responses were given while negative responses

accounted for 41 percent of the responses. Perhaps not surprisingly, given the level of

parents' education (discussed later in 6.11.3.4), the language featured in their responses

seemed to indicate knowledge about child development that was in excess of that to be

206

Chapter 6: Results

gained from the popular press directed at parents (for example women's magazines).

The wording of comments suggested that some parents were familiar with concepts of

development as they are likely to be discussed in texts directed at enhancing parenting

skills.

6.12.3.3 Responses to Questions 5,6,7 and 8

Questions Five to Eight inclusive required each parent to indicate the

bracket into which each of his or her current ages fell and the level of education

achieved by each. Results are displayed in Table 6.20.

Table 6.20

Frequencies and Percentages of Responses to Questions 5 and 6

Questions 5 and 6 (age of each parent)

Female Frequency

Percentage %

Male Frequency

Percentage %

Questions 7 and 8 (level of education achieved by each parent)

Female Frequency

Percentage %

Male Frequency

Percentage %

Under 21 years

0

0

0

0

Year 10

1

4.55

2

10

21 to 30 years

4

18

1

4.55

Year 12

3

13.5

2

10

31 to 40 years

16

73

17

77

T A F E

4

18

4

18

41 years and over

2

10

4

18

University

14

63.5

14

63.5

207

Chapter 6: Results

These findings indicate that, in keeping with national trends (ABS, 1998) all

female parents in the sample are younger than their male partners with the majority (73

percent) being aged between 31 and 40 years. While no mother was under 21 years of

age, two were older than 41 years.

6.12.3.4 The Parents in the Current Study

Responses to the questionnaires to parents showed that two thirds of them share

the following characteristics: both were tertiary-educated and employed in their

professional fields; they were familiar with appropriate developmental outcomes for

their child's age and stage of development; they possessed an understanding that play

was beneficial for children; and 'philosophical' reasons (in conjunction with others)

were cited for their choice of L D C C C . The literature discussed below supports this

writer's belief that the parents in this current study were knowledgeable in terms of both

their child's development and activities.

Tamis-LeMonda, Chen, and Bornstein (1998) examined mothers' knowledge

about play and language, finding that they had a broader knowledge of language than

play. Several reasons may account for this difference. First, in many aspects of

Western society, there seems to be a widespread focus or emphasis on language use and

ability. Also changes and increases in language skills are easily recognised and are a

readily assessable illustration of children's changing development.

Play however, is less tangible, and the content less concrete to measure or

estimate. Tamis-LeMonda et al. (1998) found in their study that mothers' understanding

and knowledge of play was less than what they knew about language. These findings

may be due to the mothers' own play experiences and how valuable these are viewed as

being. The authors suggest that society at large seems to know less about play - it is

harder to determine and the value play attracts varies greatly (Tamis-LeMonda et al.).

The findings of Tamis-LeMonda et al. (1998) are pertinent to the current study.

Although language was not in the list provided for parents' comments, 59 percent of

respondents inserted language under the heading of 'cognitive or intellectual' as an area

in which they expected their children to develop. Furthermore, parents were quite

explicit about the content of the language development they expected as they described

the gains they anticipated their children would make. Descriptions included, 'naming

body parts', 'to fit sentences together', 'formulating complete sentences' and "being able

to describe situations'. The focus on language development together with the explicit

208

Chapter 6: Results

references to it by more than half the respondents in the current study would appear to

be in keeping with the findings of Tamis-LeMonda et al. (1998).

Had the numbers of respondents been larger than the 22 parents who responded

to the questionnaire, it may have been possible to suggest stronger links with the

literature. For example, Howes et al. (1995) suggest parents of high-socio-economic

status access higher quality child-care than those of lower-socio-economic status.

Conversely two parents in the current study whose education level was below that

required for entry into professional occupations each had their children placed in

L D C C C s which did not meet the criteria considered necessary for providing quality

care. Interestingly, Holloway and Reichhart-Erickson (1989) found a positive

relationship between socially competent children, high-socio-economic status mothers

and use of high-quality L D C C C s .

In conclusion the characteristics shared by parents and discussed above could

give rise to the opinion that the results of the current study have limited generalisability

to the wider population accessing the services of L D C C C s . However, parent education

- especially that focusing on young children's play and development - is a feature of

many L D C C C parent programs. Such programs together with articles in the popular

press and those produced specifically for the 'parent' market are easily accessed by

parents and perhaps effective in raising the general awareness of the importance of play.

6.12.4 An Analysis of Responses to Staff Questionnaire

A questionnaire was completed by all staff engaged with children in the

participating L D C C C s (see Appendix Av). The aim of the questionnaire was to elicit

information regarding each staff member's knowledge of symbolic play, the value she

attributed to play in terms of influencing children's cognitive development, and her

beliefs about mixed-age free play - namely whether he or she supported the practice of

mixed-age grouping or not.

The questionnaire also had provision for staff to identify the highest level of

training achieved in early-childhood education (see Appendix Avi). Similar levels of

training and qualifications in Early-Childhood Education were grouped to form four

major categories (see Appendix Avii).

209

Chapter 6: Results

6.12.4.1 Response Rate

Fifty-eight of the 73 staff members (70 percent) completed the questionnaire.

The response rate across centres varied from 100 percent at two centres to less than 50

percent in another two centres. Response rate by centre and training are detailed in

Table 6.21.

Table 6.21

Frequency of Staff Responses by Training, Qualifications and by Centre

Category of Training and

Qualifications

Rating on N N N o n - 1 2 3 4 Environment respondents respondents

Scale U S A G

3 G 6

4 A 8

5 A 14

6 G 4

7 G 6

9 G 10

11 US 4

12 US 6

Categories of training/qualifications: Category 1 - No formal qualification Category 2 - T A P E Certificate or Diploma in Child-care Studies, Registered or

Mothercraft nurse Category 3 - Degree or Graduate Degree in Early-Childhood Education Category 4 - Other

6.12.4.2 Analysis and Discussion of Responses

The responses, to the four questions designed to elicit staff beliefs about and

attitudes toward play (Questions Five to Eight), were collated for analysis and

discussion. A more detailed account of this process together with the content of the

questions was documented in Chapter Five (5.8.4.4).

Centre Id

1

0

0

6

4

1

3

10

3

2

7

0

3

0

1

1

2

3

4

4

2

4

3

5

1

3

2

0

1

2

0

0

0

0

1

0

0

4

0

0

210

Chapter 6: Results

Positive and negative responses to these four questions were collated and are

listed in Table 6.22 in accordance with the number of staff responses and their

respective categories of training and qualifications.

Table 6.22

Frequencies and Percentages of Staff Responses by Category of Training and

Qualifications and Number of Respondents

Response P= positive N= negative P

Category of training 1

N N N N

N respondents 14 14 28 28 10 10

Question 5 Frequency 2

Percentage % 14


Percentage % 50


Percentage % 4


Percentage % 100

2

14

7

50

2

14)

14

100

18

15

28

28

18

15

28

100 100

28

90

35.5 35.5 60

10

10

10

10

40

0

0

0

100 100 100 0

5

83

3

50

6

100

6

100

1

16.5

3

50

0

0

0

0

6.12.4.3 Staff Training and Qualifications

Research supports the premise that, staff trained in Early-Childhood Education

interact with children more sensitively (Kontos & Wilcox-Herzog, 1997). Links

between the length and type of teacher training and staff behaviours with children have

211

Chapter 6: Results

also been identified (Howes, 1997). Table 6.23 illustrates the frequency of positive

responses by the most highly trained staff and the L D C C C ranked according to the

scores on the Environment Rating Scale. It is notable that where the rating awarded the

L D C C C was 'unsatisfactory' that 'category 3 trained staff were either not employed in

that centre or failed to respond to the questionnaire. (This finding was also true of one

centre rated as 'good'.) Seventy-five percent of the responses by 'category 3 trained

staff' at the remaining centres, rated as either adequate or good, were positive. While

not tested for statistical significance, such figures appear to be in the same direction as

the research cited earlier.

Table 6.23

Frequencies and Percentages of Positive Responses by Category 3 Staff by Centre

L D C C C Id Rating on environment Positive responses by rating scale category 3 staff (US, A or G) %

3 G 100

4 A 75

5 A 100

6 G #

7 G 87.5

9 G 87.5

11 US #

12 US #

US denotes unsatisfactory provision of care-giving A denotes adequate provision of care-giving G denotes good or better provision of care-giving # denotes staff with category 3 training either did not respond to questionnaire or

are not represented in the staff of that long-day child-care centre.

6.13 CONCLUSION

The results of the current investigation are presented in this chapter. A range of

data was collected for the present project. Quantitative data were statistically analysed,

and the results have been presented in the first part of this Chapter - 6.1 to 6.11.

212

Chapter 6: Results

Descriptive data were also collected and these results, mostly qualitative in nature, are

presented in part 6.12. Overall the results of the statistical analysis support the

predictions made by the present author and the descriptive data provides insight into the

environmental context in which the study was set.

The results are discussed in detail in the next chapter - Chapter Seven.

213

Chapter 7: Discussion

-/

CHAPTER 7: DISCUSSION OF RESULTS

7.1 INTRODUCTION

This chapter contains a detailed discussion of the results of the current enquiry

set out in the previous chapter and in Appendix I.

Answers were provided, in part, to the major hypothesis and the related

questions. The results for the major hypothesis and the first research question are

examined first, followed by those relevant to each of the two additional research

questions.

Each result is discussed first in the context of this study and secondly, in terms

of the literature reviewed in Chapter T w o . Additionally the results are explained from

the two major theoretical perspectives underpinning this research project (Piaget's

theory of cognitive development and Vygotsky's socio-cultural theory of cognitive

development). The perspectives of other theorists whose theoretical approaches were

explored in Chapter Three are referred to where relevant.

The results supported only some of the predictions made - those inconsistent

with the expectations are presented and discussed later in this chapter (7.18).

Implications arising from this discussion are detailed towards the end of this chapter in

7.20.

7.2 THE HYPOTHESIS AND ASSOCIATED RESEARCH QUESTIONS

As described earlier, the major hypothesis is that toddlers engaged in free play

with older peers (in either mixed-age play or dyad play) would exhibit symbolic-play

skills in each of the four dimensions of symbolic play identified by Westby (1991) -

'decontexutalisation', 'thematic content', 'organisation of themes' and 'self-other

relationships'- more frequently and at a higher level, than when they are engaged in


In conjunction with the major hypothesis, three research questions were also

considered.

214


7.3 THE THEORETICAL BASES FOR THIS RESEARCH

As elaborated upon in Chapter Three, this project was developed in keeping with

elements of the theories of the following theorists: Piaget, Vygotsky, Bandura, Erikson,

Bruner, and Bronfenbrenner. Each theory proved useful in establishing the research

project and contributing to the interpretation of the results of the study. Reference to the

theories of Piaget and Vygotsky ensured results were considered from both a cognitive

and socio-cultural approach. Piaget and Vygotsky each viewed symbolic play, the

central concern of this project, as an essential element in the lives of young children and

their subsequent development. These two cognitive theories - Piaget's theory of

cognitive development and Vygotsky's socio-cultural theory - largely underpin the

research method and the major hypothesis. Therefore Piagetian and Vygotskian

perspectives will each be adopted to interpret the results related to the questions

contained in the major hypothesis and the additional related questions. Reference to the

work of McCune-Nicolich (1977, 1981) and the neo-Piagetians - Halford (1993), Case

(1985,1991), Fischer & Farrar (1987) - extend the cognitive perspective.

Allusion to the psychosocial stance of Erikson (1963), and the cognitive-social

learning theory approach proposed by Bandura (1986, 1989), broadened the

developmental interpretation of the results. The ethological and ecological perspectives

of Bruner (1972, 1973, 1977) and Bronfenbrenner (1979, 1993) assisted the researcher

to set the results in a meaningful context.

7.4 STATISTICAL ANALYSIS OF DATA

As described in Chapter Five (5.11.1), four statistical tests - Friedman's Two-

W a y Analysis of Variance by Ranks, Wilcoxon's Matched-Pairs Signed-Ranks Test,

The Kruskal-Wallis One-Way Analysis of Variance by Ranks and Logistic Regression

Analysis - were used in this investigation. Results of these tests supported, in part, the

direction predicted in the major hypothesis. These four tests also helped answer three

research questions additional to the major hypothesis. The results will n ow each be

discussed in conjunction with the relevant questions.

215


7.5 RESULTS FOR THE MAJOR HYPOTHESIS AND THE FIRST RESEARCH

QUESTION

As foreshadowed earlier in this chapter (7.2) the focus of the major hypothesis

was the frequency of participants' symbolic-play events and the levels of participants'

symbolic play (in the dimensions of 'decontextualisation', 'thematic content',

'organisation of content' and 'self-other relationships') when they were engaged in free

play with their older preschool-aged peers. The first research question involved the

identification of the most influential condition - whether mixed-age play or dyad play.

7.5.1 Frequency of Symbolic-Play Events

As shown in Chapter Six (6.5) the prediction, that toddlers when playing with

older peers in mixed-age play and dyad play, would exhibit symbolic play more

frequently than when playing with their same-age peers, was supported. Friedman's

Rank test showed that in Condition Three (dyad play), when participants were engaged

in play with a preschool-aged peer, a significant difference existed between the means

for the frequency of symbolic-play events. The frequency of symbolic-play behaviours

was higher in dyad play (Condition Three) than in either Condition One (same-age play)

or Condition T w o (mixed-age play).

In summary, most of the symbolic-play events occurred for all participants in

dyad play (Condition Three). However, a relationship between birth order and

symbolic-play activity was evident as first-born participants showed the most events of

symbolic play in mixed-age play (Condition Two). (This result is discussed later in this

chapter in part 7.10.)

In conclusion, predictions concerning the frequency of the participants'

symbolic-play events were supported. Evidence cited above shows that both mixed-age

play and dyad play favourably influence the frequency of symbolic-play events.

7.5.2 The Highest Levels of Symbolic Play

The prediction that toddlers, when playing with preschool-aged peers, would

exhibit symbolic-play activity at higher levels was supported. Friedman's Rank test

showed significant results in Condition Three (dyad play) for Dimension Three

('organisation of themes') and Dimension Four ('self-other relationships'). N o

significant differences between the means for Dimension One ('decontextualisation')

and Dimension T w o ('thematic content') in any of the three conditions were found.

216


7.5.3 The Complexity of Symbolic-Play Activity

Complexity was defined as symbolic play that was either, 'greater than' or 'less

than or equal to' that expected in relation to the participant's age (according to Westby's

1991 scale). Logistic Regression Analysis was used to find first, the relationship

between individual ages and the 'likelihood' of participants showing complexity of

symbolic play and secondly, to identify from a range of variables those useful in

explaining the 'likelihood' (or odds ratio) for complexity of symbolic play to occur.

The relationship between individual age and complexity of symbolic play was

statistically significant in Condition Three (dyad play) in three dimensions; Dimension

One ('decontextualisation'); Dimension Three ('organisation of themes'); and

Dimension Four ('self-other relationships').

Additionally, Condition Three (dyad play) was one of two variables found to be

influential for complexity of toddler participants' symbolic play to occur in two

dimensions: Dimension T w o -'thematic content'; and Dimension Four - 'self-other

relationships'. (Other variables found to be influential on the complexity of

participants' symbolic play are considered later in this chapter in part 7.11.)

7.6 DISCUSSION FOR THE MAJOR HYPOTHESIS AND QUESTION ONE

Westby's Scale for Assessing Children's Pretend Play was based on " 8 0 % of

middle-class preschoolers performing the play and language behaviours at each level"

(1991, p. 133). Westby continued, " it was c o m m o n for many middle-class children to

exhibit the behaviours identified on the scale at somewhat younger ages, particularly in

the earlier stages" (1991, p. 133). Westby's statement is interpreted as indicating that

children from middle-class backgrounds - especially the younger ones - were likely to

engage in symbolic-play behaviours earlier than was expected. The findings of the

present study, in which the parents of the participants were largely tertiary-educated and

employed in professional capacities, lend weight to the second of Westby's statements

cited above and confirm the soundness of the participant sample used in this project.

Clearly age, development and experience are factors contributing to young children's

competencies in symbolic-play activity. Westby's (1991) claim may perhaps not be

surprising given that research findings (for example those of de Lacey, Walker, &

Nurcombe, 1999) have shown that middle-class preschoolers are also ahead of others on

measures of verbal IQ.

217


As reported above toddlers engaged in play with a familiar preschool-aged peer

in Condition Three (dyad play) exhibited higher-level symbolic-play skills than they

exhibited in Condition One (same-age play) or Condition T w o (mixed-age play). The

higher level skills exhibited in Condition Three (dyad play) were in two dimensions -

'organisation of themes' and 'self-other relationships' - the third and fourth dimensions

of symbolic play referred to in the Westby assessment scale (1991).

'Organisation of themes' is the component of symbolic play, which indicates the

manner in which a child links together (or sequences) actions in play. As the skill of

organisation increases, so the resultant play becomes more integrated and coherent.

Initially, in very young children, symbolic-play actions are short, described by Westby

as "isolated schemas (single-pretend actions)" (1991, p. 136). The participants in the

present study showed a higher level of organisation in that their symbolic-play actions

were linked or sequenced in a way above that expected at their age.

Dimension Four ('self-other relationships') was the second dimension in which

participants displayed a higher level of competency than would be expected at their

ages. This enhanced level of competency was only exhibited in Condition Three (dyad

play) and not in either of the other two conditions - Condition One (same-age play) or

Condition T w o (mixed-age play).

Westby (1991) described this dimension of symbolic play ('self-other

relationships') as one that "frees symbolic actions from children's own bodies" (1991,

p. 141). Rubin, Fein, and Vandenberg (1983) in keeping with other researchers

recognise that this behaviour follows a developmental sequence. Initially symbolic

actions are focused on the self - for example the child feeds him or herself or drinks

from a baby's feeding bottle. Westby terms this behaviour as the child engaging in

"auto-symbolic or self-representational play" (1991, p. 136).

Such activity is described elsewhere in the literature as 'self-referenced'

behaviour as the child is the 'active agent' (Rubin et al., 1983). Later, in the

developmental process, the child is able to engage in 'other-referenced' symbolic-play

behaviour whereby an object (for example a doll) or an adult may be the "passive

recipient or the object of the child's action" (Rubin et al., 1983, p. 718). The highest

level in 'self-other relationships' (in the range of this study) is when the child is able to

manipulate an object (for example a doll or some other appropriate item) as if it were

the active agent. The taking on of others' roles in coordinated, collaborative play, the

218


most sophisticated level of Dimension Four, occurs at approximately 60 months

(Westby, 1991) and falls well outside the expected findings of this project.

Statistical analysis indicated that, on average, the participants in this study, in

Condition Three (dyad play), exhibited Dimension Four behaviour ('self-other

relationships') above that expected for their age. (Further theoretical discussion can be

found in parts 7.8, 7.15 and 7.16 of this chapter.)

7.7 LITERATURE SUPPORTING THE CURRENT FINDINGS

In the literature reviewed earlier (Chapter T w o , 2.2), a number of other

investigators interested in the influence of older peers on younger children's play

observed children in either group or dyad conditions. The findings, arising from each

method and pertinent to this study, are summarised below.

7.7.1 Research Featuring the Dyadic Model

In the current project significant results were most evident in the third condition

only, dyad play. Lougee, Gruenich, and Hartup (1977), Howes and Farver (1987), and

Lyytinen (1991, 1995) focused specifically on dyadic situations in assessing the benefits

for the younger peer in mixed-age free play.

In those studies cited above the younger children in the dyad pair were found to:

increase their attention to the play activity (Lougee, et al., 1977); exhibit more

cooperative social pretend play and imitation (Howes & Farver, 1987); and engage in

more play activity and pretend play at a more complex level (Lytinnen, 1995).

The dyad-play situation in the current enquiry features a number of specific

conditions that are unlikely to operate in other mixed-age play sessions. First, and

perhaps of paramount importance, the toddler participant was with a self-chosen,

familiar peer - one w h o m he or she favoured as a play partner in the mixed-age play

situation. Secondly, the older peer was requested to help the younger child play with

the toys and materials. This request was made with the expectation that it might

encourage the older peer to engage in tutoring or assisting the younger child in play. In

addition, there were adequate materials available for both players - unlike group-play

sessions where there is often a need to compete for materials. The dyad-play condition

took place a little to the side of the main activity area so interruptions were rare. This

positioning also enabled optimal visual cues to assist the toddlers' imitation.

219


In conclusion the significant results - in terms of the frequency of symbolic-play

events, higher levels and complexity of symbolic activity - shown in the dyad-play

condition in the present project are in keeping with those of previous researchers. The

results add support to the notion that the dyad is a useful medium for promoting

younger children's enhanced play activity.

7.7.2 Research Featuring the Group Model

A second body of research (also reviewed in Chapter Two, 2.2) supports the

notion that younger children, who attend L D C C C s , benefit from placement in mixed-

age groups for free play (Goldman, 1981; Mounts & Roopnarine, 1987; Bailey,

McWilliam, Ware, & Burchinal 1993; Bailey, Burchinal, & McWilliam 1993). The

younger peers were seen to benefit in the following ways: spending less time in parallel

play (Goldman, 1981); showing an increase in the cognitive content of their play

(Mounts & Roopnarine, 1987); spending most of their time playing with older children

(Bailey, McWilliam, et al., 1993); and achieving higher scores in all developmental

domains (Bailey, Burchinal, et al., 1993).

Mixed-age grouping found support in the results of two more studies - those of

Rothstein-Fisch and Howes (1988) and Dunn, Kontos, and Potter (1996) - each

conducted in Family Day Care settings where mixed-age grouping is the standard.

Toddlers in the Rothstein-Fisch and Howes (1988) investigation were more active and

aroused in complex interaction with their 2- to 3- year old peers while the participants in

the Dunn et al. (1996) study showed more complex cognitive play behaviour with their

older peers.

In the current study first-born participants showed significantly more frequent

symbolic-play events and higher levels of symbolic play in mixed-age play (Condition

Two). (An explanation of these results follows in part 7.10). Mixed-age play, while

indicated by these results as certainly beneficial for young children, lacks the specific

features of the dyad-play condition as described earlier. In addition, mixed-age play in

Australia (perhaps in contrast to the early-childhood settings mentioned in the literature)

has largely been introduced due to pragmatic, rather than theoretical, reasons.

Consequently, mixed-age play sessions may be more ad hoc than those featured in the

research cited.

Clearly the present study did not attempt to replicate each of the investigations

referred to above, but the researcher's expectations were that results would indicate

220


some similarities in direction. (These expectations gain further support from the

descriptive statistics in Chapter Six, 6.5, and the accompanying tables.)

7.8 THE RESULTS FROM A VYGOTSKIAN THEORETICAL PERSPECTIVE

A socio-cultural, Vygotskian perspective is useful when explaining the results

thus far. Significant differences were found between the means for the participants'

highest levels of two symbolic-play behaviours, Dimension Three ('organisation of

themes') and Dimension Four ('self-other relationships') and the frequencies of

symbolic activity in Condition Three (dyad play). Statistically significant results were

found for the relationship between individual age and complexity of symbolic play in

Condition Three (dyad play) in three dimensions; Dimension One

('decontextualisation'); Dimension Three ('organisation of themes'); and Dimension

Four ('self-other relationships'). Additionally, Condition Three (dyad play) was found

to be a useful variable to explain the 'likelihood' or 'odds ratio' of participants showing

complexity of symbolic play in Dimension T w o ('thematic content') and Dimension

Four ('self-other relationships').

These findings accord with two important Vygotskian premises. First, Vygotsky

(1966) appeared to consider play itself as contributing more to cognitive development

than did Piaget. He described play as giving a child "much keener experiences" than

purely pleasure (1966, p. 6). In support of that premise, Vygotsky believed that when a

child was engaged in play he or she was able to behave in a manner that is a little ahead

of his or her actual developmental stage.

The second premise is the 'zone of proximal development' (ZPD), a Vygotskian

construct which proposes that an individual, in the company of a more accomplished

other, achieves more success than was possible independently. Vygotsky considered

that play in itself may act as a Z P D - permitting children to 'self-scaffold' - and support

themselves in performing or functioning at a higher level than is expected for their age

or stage of development (1978, p. 103).

It is likely that each of the premises cited above was an active component of the

toddler-participants' experience in the current study. However, an older preschool-aged

partner - likely to be more skilled in the ways of symbolic play than the participant -

was a feature of the dyad situation - Condition Three, the condition in which the

significant results currently being discussed occurred. It seems likely that this partner, a

221


'more-accomplished other' was effective at invoking a Z P D in emerging symbolic play

for the toddler participant.

The third result, that significantly more events or instances of symbolic-activity

occurred in dyad play, m a y be explained from a Vygotskian perspective as an outcome

of "guided participation" (Rogoff, 1990, p. 8). This is best defined as the "process and

system of involvement of individuals with others, as they communicate and engage in

shared endeavors," (Rogoff, Mistry, Goncu, & Mosier, 1993, p. 6). Embedded in this

notion is the belief that, while children are active in their o w n learning, interaction with

others (from their environment) is essential to impart, through social interaction,

valuable and essential cultural knowledge (Rogoff, 1990). Rogoff suggests children

may act as "important cognitive facilitators" for one another (1990, p. 183).

Condition Three (dyad play) fulfils this requirement in a number of ways. First,

the toddler participant was certainly active, engaging with both the materials provided

and the 'other' - the older preschool-aged peer. Secondly, the older peer partner, the

more knowledgeable and skilled other, was usually more than willing to guide the

toddler in the 'ways' of symbolic activity. Essentially the older peer in the dyad

situation presented toddlers with guidance and at times challenges, simultaneously

illustrating the 'rules of the activity' that in turn reflect the rules of society. Finally the

play partner was from the toddlers' "immediate social context," (Rogoff, 1990, p. viii)

The participants' enhanced performance in Dimension Three ('organisation of

themes') and Dimension Four ('self-other relationships'), together with the frequency of

symbolic activity shown in this study, could each have been brought about through

either or both of the premises explored above.

Berk (1994) describes the central concept of Vygotsky's theory as thus: "all

uniquely human, higher forms of mental activity are jointly constructed and transferred

to children through dialogues with other people" (1994, p. 30). Condition Three, dyad

free play, provided a forum for such dialogue.

7.9 RESULTS FOR RESEARCH QUESTION TWO

The question posed was whether first-born participants would first, exhibit

symbolic-play events more frequently than 'later-born' participants and, secondly,

engage in higher levels of symbolic play than later-born participants.

222


7.9.1 The Relationship Between Birth Order and the Frequency of Symbolic-Play

Events

The relationship between the participants' birth order and frequency of

symbolic-play events was examined using the Kruskal-Wallis test. While no significant

results were found for the relationship between birth-order and the frequency of

symbolic-play events a p value of .061 approaching significance was recorded for first-

bom participants in Condition T w o (mixed-age play).

7.9.2 The Relationship Between Birth Order and the Level of Symbolic Play

The Kruskal-Wallis statistical test showed that significantly higher differences

between the means of the highest levels of each of the four symbolic-play dimensions

were found in mixed-age play only (Condition T w o ) for first-born participants.

Significant results were shown for: Dimension One ('decontextualisation'); Dimension

T w o ('thematic content'); Dimension Three ('organisation of themes'); and Dimension

Four ('self-other relationships'). These outcomes supported the researcher's

expectation that first-born participants would exhibit higher-level symbolic-play skills

than later-born participants.

The results detailed above only partially accord with the predictions. It was

expected that more frequent events and higher levels of symbolic play would occur in

both mixed-age play (Condition T w o ) and dyad play (Condition Three). There were no

significant results for the relationship between birth order and frequency of symbolic-

play events and higher levels of symbolic play in Condition Three. (The last results,

inconsistent with the predictions, are discussed in part 7.18 of this chapter.)

7.10 DISCUSSION: THE RELATIONSHIP BETWEEN BIRTH ORDER AND

SYMBOLIC PLAY

7.10.1 First-Borns

As noted earlier, an additional research question was whether first-born toddler

participants would exhibit more frequent symbolic-play acts (or events) and higher level

symbolic play than 'later-born' toddlers.

The literature informing this question is reviewed in Chapter T w o , (2.3 & 2.4)

comes from two sources. First, to be considered is the research identifying parental

(usually maternal) influences on children's symbolic play and secondly, investigations

223


into the influence of older siblings on their younger siblings' activity. The question

posed arose from the recognition that parental (most often maternal) input is influential

in very young children's emerging symbolic play. The literature reviewed, for example,

Dunn and Wooding (1977), Slade (1987a, 1987b), Fiese (1990) and Bornstein, Haynes,

O'Reilly, and Painter (1996), supported the prediction that first-borns would exhibit

both more frequent and higher levels of symbolic play. Information contained in a

review of the research considering outcomes for 'only' children was also useful (Falbo

& Polit, 1986).

It may be recalled that there were no significant differences between the means

for the frequency of symbolic-play events for either first-born or later-born participants.

However, there was a trend close to statistical significance, (p = .061), for first-born

participants to exhibit symbolic play more frequently than later-born participants, in

Condition T w o (mixed-age play).

Toddlers who were first-borns showed significantly higher levels of all four

dimensions of symbolic-play behaviour in Condition T w o (mixed-age play) than in

either Conditions One or Three (same-age play or dyad play).

The experience of first-borns is close to that of 'only' children (Falbo & Polit,

1986). Family relationships, both immediate and extended, are likely to be closer, and

more time is spent in close proximity with family members (often the mother) for w h o m

perhaps a degree of novelty exists in terms of discovering, monitoring and taking a

facilitative role in the child's developmental progress (Falbo & Polit, 1986, p. 183). It

maybe that, in the light of the results for first-borns in this study, those participants

came to the L D C C C equipped with skills, experience and expertise as a result of their

first-born status.

7.10.1.1 Relevant Literature

The role of parents (more especially mothers) in the development of very young

children's symbolic play has come under regular and close scrutiny from a number of

researchers (O'Connell & Bretherton, 1984; Haight & Miller, 1992; Damast, Tamis-

LeMonda, & Bornstein, 1996). Consensus exists that mothers are instrumental in aiding

and promoting the development of symbolic activity.

The involvement of mothers with their children in play has been found to be

associated with increases in five variables: the amount and diversity of 20-month olds'

symbolic play (O'Connell & Bretherton, 1984; Bornstein et al., 1996); both the level

224


and the duration of increase the complexity of symbolic play of 15 to 24 months olds -

especially if they engaged with their child in a modelling task - Fiese (1990); and the

completion of symbolic-play sequences when engaged in 'joint attention' with their 18

to 24 month olds (Dunn & Wooding, 1977).

In addition, mothers have been found to adjust their responses to their 21-month-

old children's play (Damast et al., 1996) and purposefully direct pretend play toward

their 12-month-olds (Haight & Miller, 1992).

The research involving fathers and their children's symbolic play is scarce but

evidence suggests that the direction is the same. Findings included: an increase in

participants' symbolic play with both age and their fathers' involvement (Farver &

Wimbarti, 1995); and reciprocal symbolic-play responses between both parents and

their children (Haight, Parke, & Black, 1997).

(Importantly for the current discussion, participants in the studies of Damast et

al. [1996], Bornstein et al. [1996] and Haight et al. [1997] were first-borns.) In

conclusion the research referred to confirms that partnerships with parents enhance

young children's symbolic play along a number of parameters.

As will be noted a little later in this chapter, (7.11.1), an explanatory variable for

the participants' higher levels of the symbolic play behaviour, 'organisation of themes'

(Dimension Three), was their attendance on four or more days' at the L D C C C . (As the

toddlers in the current study attended day-care for about six to eight hours a day, four or

more days attendance was the equivalent of between 24 and 32 hours a week.) It may

be likely that such attendance provides the opportunities for children to further practise

the symbolic-play skills they acquired in their play with their mothers.

It m a y also be that those toddlers attending four or more days a week are from

higher socio-economic backgrounds than participants attending less than four days a

week. (This m a y be an outcome of the 1997 Federal Government reduction in child

care funding resulting in increased fees and a decrease in the child-care rebate.)

7.10.2 Later-Borns

It is possible to view the relationship between birth-order and symbolic play

from another perspective, in that participants with older siblings m a y be well acquainted

with and well practised at symbolic play through activity with their siblings. Such

experience m a y render these toddlers (approximately half the participants in this current

study) unreceptive when opportunities for symbolic play arise in the L D C C context in

225


that other activities, engaged in less frequently, seem more attractive options. Support

for this proposal - that older siblings support the symbolic play of their younger siblings

- is found in the literature described below.

7.10.2.1 Relevant Literature

It will be recalled that in Chapter T w o (2.7) the efficacy of older siblings in the

support of their younger siblings' symbolic or pretend play is illustrated in a number of

studies. Benefits, for the younger sibling, arising from sibling dyads included: symbolic

play featuring daily routines and domestic activities (Dunn & Dale, 1984); more

frequent and more complex symbolic play (Dale, 1989; Youngblade & Dunn, 1995);

support and guidance in the use of play materials (Lamb, 1978); relationships existing

between their active and high-level play and language development; their play attracting

responses and/or reactions; and receiving communicative behaviours (Teti, Bond, &

Gibbs, 1988)

In addition to the results cited above, Dunn and Dale (1984) found the younger

siblings imitated their older siblings' pretend actions and comments and L a m b (1978)

noted that the younger siblings monitored the older siblings' behaviour. It is likely that,

the toddlers' imitation of their older siblings behaviour is a component in their

acquisition of symbolic-play skills. In the absence of older siblings it may be that in the

L D C C C environment toddlers imitate their older peers.

The work of the following researchers supports the proposition that toddlers

imitate the behaviour of older peers: Abramovitch, Corter, and Lando (1979);

Abramovitch, Corter, and Pepler (1980); and Samuels (1980). (Each of their studies is

discussed in more detail in the Literature Review - Chapter T w o , 2.4.)

7.11 RESULTS FOR RESEARCH QUESTION THREE

Logistic Regression Analysis was used to identify the explanatory variables (or

influences) on the participants' complexity of symbolic play. Complexity was viewed

as either 'greater than' or 'less than or equal to', the expectations (in terms of age)

identified in Westby's (1991) Scale for Assessing Children's Pretend Play. Results of

the analysis indicated the following three variables as 'useful' predictors: unsatisfactory

provision of play materials; an attendance of four or more days a week at L D C C C ; and

Condition Three (dyad play). (It is noted that each of the ascribed influential variables

226


may well be a reflection of other unknown factors and not necessarily as 'causal' - for

example, attendance patterns of four or more days may indicate that parents are very

busy.)

A discussion of the dyad condition in relation to context of the study and the

literature reviewed was presented earlier in this chapter (7.6 & 7.7.1) and Vygotskian

theoretical perspectives of the dyad condition were discussed in 7.8. (Piagetian

theoretical perspectives and those of other theorists are explored later in parts 7.15 and

7.16 of this chapter.) The remaining two explanatory variables, participants' attendance

at L D C C and unsatisfactory provision of play materials, are explored below.

7.11.1 The Participants' Weekly Attendance at LDCC

As shown in the results, participants' attendance of four or more days a week at

L D C C is an explanatory variable for the 'likelihood' (chance or odds ratio) of the

symbolic play being more complex than that expected for their age.

Evidence in support of children's attendance at L D C C is found in the literature

reviewed in Chapter T w o (2.10). Positive outcomes for participants who attended

L D C C C s included engagement in both more cooperative play (Aureli & Procacci, 1992)

and more complex symbolic play (Aureli & Collechia, 1992).

Importantly children attending L D C C C s on a regular basis experience a shared

environment (both physical and social) with a group of peers with w h o m they are likely

to be well acquainted. Matthews (1977b) found an effect for peer familiarity on the

fantasy play of 4-year olds while Rubenstein and Howes (1976) observed 19-month olds

engaged in higher level play in the presence of a familiar, same-age peer. It is likely

that in the L D C C C environment the toys, materials and equipment provided for

children's play are repeated from time to time. Perhaps this repetition of play materials

encourages children also to repeat their ensuing play activities.

In conclusion, it seems that participants attending L D C C C for four or more days

weekly, are likely to exhibit enhanced play skills. Children may experience practice, in

both the provisions for play and their peers' play activity, to a greater degree than those

participants w h o attended for less than four days a week.

7.11.2 The Unsatisfactory Provision of Play Materials

The expectation was that the satisfactory provision of play materials was likely

to be a positive influence upon the results. Conversely, when the complexity of

227


symbolic play was analysed, a variable indicated as 'useful' in explaining the

'likelihood' (or odds ratio) of participants exceeding expected levels of symbolic-play

activity was the unsatisfactory provision of play materials.

It could be that, in an effort to support young children's developing symbolic

play, an abundant or an over-supply of realistic, life-like props exists in the play

environment. Such provision m a y reduce the necessity for children to 'transform'

objects into those required or to use one object to represent another. Tegano,

Lookabaugh, May, and Burdette (1991) found that an increase in children's constructive

play (often a component of symbolic play) was not related to the level of planned

structure by the teacher. Interestingly, they found that, as children imposed their own

structure, so their 'nonplay' decreased, on the contrary, nonplay increased with teacher

imposed structure.

Steele and Hrncir (1985) hypothesised that toys with 'high prototypicality' (their

term for 'high realism' or 'life-like') would prompt a higher frequency of symbolic play

from participants than props considered 'low prototypicality'. Results contradicted this

hypothesis leading Steele and Hrncir to consider that possibly the similarities between

the two groups of props used in the study were such that the children failed to perceive

their differences. Perhaps, for the present discussion, the more significant of their

findings was that the adults in proposing play suggestions using props fitting the low

prototypicality description supported the children's transformations.

In part 7.17.2 of this chapter the features of one participating L D C C C are

explored in some detail supplying anecdotal support for some of the considerations

explored above.

7.12 CONCLUSIONS DRAWN FROM THE RESULTS DISCUSSED THUS FAR

The predictions made in this current enquiry are in part supported by the results

discussed above in 7.5, 7.9 and 7.11. Additionally the findings indicate directional

trends described in the predictions. (Results which are inconsistent with the predictions

are discussed later in this chapter in 7.18.)

228


7.13 ADDITIONAL RESULTS ARISING FROM THE STATISTICAL

ANALYSES

7.13.1 The Most-Frequently Occurring Levels of Symbolic Play

After the major hypothesis and related questions were formulated the notion of a

participants' most-frequently occurring levels of symbolic play was recognised as likely

to provide useful information in terms of the participants' potentials for highest level of

symbolic play.

The Friedman's Rank test was used to establish whether or not differences

existed between the means of the participants' most-frequently occurring levels of each

of the four symbolic-play dimensions ('decontextualisation', 'thematic content',

'organisation of themes' and 'self-other relationships') exhibited by the participants in

each condition (same-age play, mixed-age play and dyad play). The test showed that

there were no significant differences between these means in any dimension in any of

the three conditions.

7.13.2 The Relationship Between Birth Order and the Most-Frequently Occurring


The relationship between birth order and the most-frequently occurring levels of

symbolic play was then examined using the Kruskal-Wallis test. Results showed that

significant differences between the means for the most-frequently occurring levels of

symbolic play were found for first-born participants in Condition T w o (mixed-age play)

only. Significant differences were found in three dimensions: Dimension T w o

('thematic content'); Dimension Three ('organisation of themes'); and Dimension Four

('self-other relationships'). In addition a p value of .067, which approaches

significance, was shown for Dimension One ('decontextualisation').

7.13.3 The Relationship Between Age and Symbolic Play

It was thought useful to examine the relationship between both the participants'

age group and their individual age and symbolic play. First, participants were grouped

in age largely in accordance with Westby's (1991) Scale for Assessing Children's

Pretend Play and the relationship between age group and symbolic play was examined

using the Kruskal-Wallis test.

This test showed a relationship between age group and higher levels of symbolic

play for participants aged 20 to 22 months. Participants in this age group scored

229


consistently lower than participants in the other three age groups. Results were in two

dimensions in Condition One (same-age play) as follows: Dimension One

('decontextualisation'); Dimension T w o ('thematic content'); and Dimension Four

('self-other relationships'). In addition two p values - .074 for Dimension Three

('organisation of themes') and .056 for Dimension Four ('self-other relationships') -

approach significance. (An explanation for these findings - inconsistent with

predictions - is presented in the following section, 7.14.)

Logistic Regression Analysis was used to test for the relationship between the

participants' individual age and the complexity of symbolic play. Results showed that

the chance of exceeding the 'expected' level depended on the age of the child (note each

child's chronological age was entered - not by group). A relationship was found

between the participants' ages and the complexity of their symbolic play. This

relationship was shown to be most strongly in the youngest age group - participants

aged 17 to 19 months. In Condition One (same-age play) age was shown as the

predicting factor for participants to go beyond or exceed the expected level of symbolic

activity in Dimensions T w o and Four ('thematic content' and 'self-other relationships').

Additionally a p value of .06 for Dimension One ('decontextualisation') approaches

significance. Development in each of these three symbolic-play behaviours is closely

tied to children's maturation, experience and cognitive development. These factors are

discussed later in this chapter (7.15 and 7.16) when theoretical perspectives are

explored.

In Condition T w o (mixed-age play) the relationship between age and the

complexity of symbolic play is statistically significant for Dimension One only. Finally

in Condition Three (dyad play) the relationship is highly significant in three dimensions

as follows: Dimension One ('decontextualisation'); Dimension Three ('organisation of

themes'); and Dimension Four ('self-other relationships'). This effect appears most

strongly in the youngest age group (17-19 months). (Results were displayed in Chapter

6, Table 6.14).

Not surprisingly it is the older age group (27 to 31 months) w h o are most likely

to exceed expectations in Dimension One ('decontextualisation'). The process of

'decontextualisation' it is recalled, involves the lessening of a dependency on life-like

(or high-realism) props for constructing symbolic play.

Logistic Regression Analysis showed that Condition Three (dyad play) was the

condition in which the relationship between age and the likelihood of participants

230


exceeding the expected level of symbolic activity was most significant. Significant

effects were found in two dimensions of symbolic play ('decontexualisation' and 'self-

other relationships'). (As mentioned earlier, variables likely to have influenced these

results are discussed later in part 7.17 of this chapter.)

7.14 A COMMENT ON THE RESULTS FOR PARTICIPANTS AGED 20 TO 22

MONTHS

In the current study six participants from four different L D C C C s comprise the

20- to 22-month-old group. Though the following trends are noted with caution because

of the small number of participants, nevertheless, they are interesting and probably

worthy of future investigation.

W h e n the relationship between age group and frequency of symbolic-play events

and levels of symbolic play - in the four dimensions identified by Westby (1991) - was

tested the participants aged 20 to 22 months featured in a number of ways. First, they

exhibited less-frequent events of symbolic play than the participants in the other three

age groups in two conditions - same-age play (Condition One) and dyad play

(Condition Three).

Secondly, these toddlers exhibited significantly lower levels of symbolic play

than children in the other three age groups - again in same-age play (Condition One) in

the three dimensions of 'decontextualisation', 'thematic content' and 'self-other

relationships'. Also, the p value for the dimension 'organisation of themes' was

approaching significance.

Finally when the most-frequently occurring levels of symbolic play were

examined this age group, showed significantly less symbolic activity than members of

the remaining three age groups. These results were shown in same-age play (Condition

One) in two dimensions of symbolic play ('decontextualisation' and 'self-other

relationships') - with the third ('organisation of themes') and fourth ('self-other

relationships') approaching significance.

Participants aged 20 to 22 months are, compared to the younger children likely

to be in their same-age group in L D C C C s , somewhat more independent and less

requiring of support, direction or assistance. It is possible that their developing social,

language, physical and cognitive abilities render them to some extent 'invisible' to staff

caring for them within a group of children - likely to be mostly younger than 20 months

231


who need more routine attention and support than the 20- to 22- month-old. The

significant findings for the relationship between age group and symbolic play were all

in same-age play (Condition One) during which time, other than staff, it is likely that no

'others' were available to fulfil the role of the 'older, more-accomplished other' for

interaction with these the 'older' members of the group.

7.15 THE RESULTS FROM A PIAGETIAN PERSPECTIVE

As discussed earlier in this chapter (7.5.2) participants showed higher levels of

symbolic-play activity in two dimensions of symbolic play, Dimension Three

('organisation of themes') and Dimension Four ('self-other relationships'), in dyad play

(Condition Three).

7.15.1 Symbolic Play

Piaget (1954) proposed that symbolic-play activity requires the child to behave

in a manner that is representational and not related to the current reality. It is useful to

consider the range of symbolic activity as understood by Piaget. H e described the

beginning of symbolic activity as "the union of two conditions - applications of the

schema to inadequate objects and evocation for pleasure" (1962, p. 97). Consolidated

symbolic activity he described as involving actions "applied to new and inadequate

objects . . . carried out with strict attention to detail although they are entirely make-

believe" (1962, p. 101).

It is possible that, during the very early stages of symbolic-play development,

such activities m a y gain more validity if they are demonstrated and engaged in by an

older peer in a free-play situation. Although Piaget considered that children

collaborated in play only from approximately four years of age he, postulated, that

"collaboration between equals is first and foremost collaboration in play" (1959, p. 62).

Piaget's use of the term 'equals' suggests a shared orientation, context and knowledge

as prerequisites to successful collaboration. Interestingly he also referred to

collaboration (on this occasion with reference to children's conversations) as "help

given by an older to a younger child" (1959, p. 62).

Researchers in the Piagetian tradition, Stambak and Sinclair (1993) remark how,

in earlier work with their colleagues (Stambak et al., 1983 - published in French only) -

they observed toddlers engaged in play with others often "center their attention on the

232


same idea of what can be done with the objects" (1993, p. ix). (This notion perhaps

contains elements of the two concepts of "joint attention" [Tomasello, 1995, p. 107] and

"joint involvement episodes" [Schaffer, 1996, p. 252].) Stambak and Sinclair noted that

these very young children would adopt another's ideas and actions and then elaborate

upon them (1993).

These observations and the current findings lend support to Piaget's proposition

(concerned with imitation and causality) that "it is very probable . . . that contact with

persons plays an essential role in the processes of objectification and externalization"

(1954, p. 252). In this current study, Condition Three (dyad play) in which the toddler

participant shared play materials with a familiar 'self-chosen' preschool-aged peer (a

peer identified as one with w h o m the toddler engaged in mixed-age play - Condition

Two) supports this notion of benefit from contact with another (or others) in a shared

framework

7.15.1.1 'Decontextualisation' (Object Substitution): Dimension One

With the exception of those first-born toddler participants who showed

statistically significantly higher levels of 'decontextualisation' symbolic behaviour in

Condition T w o (mixed-age play) and in Condition Three (dyad play) where activity in

'decontextualisation' approached significance, toddler participants did not show

enhanced skills in this dimension.

As stated earlier, in Westby's (1991) scale, Level One 'decontextualisation'

remains applicable until Level Six - that is, it remains unchanged while other

dimensions increase in complexity throughout Levels One to Six. While anecdotally

there were 45 instances of 'decontextualisation' behaviour coded as Level Six, these

were too few to be statistically significant. (The distribution of instances, or events, of

'decontextualisation' are as follows: 18 instances in Condition One - same-age play; 22

in Condition T w o - mixed-age play and five in Condition Three - dyad play.)

Results showed that it is the older age group (27 to 31 months) w h o are most

likely to exceed expectations in Dimension One ('decontextualisation'). The process of

'decontextualisation' involves the lessening of a dependency on life-like (or high-

realism) props for constructing symbolic play. In accordance with Piagetian theory,

McCune-Nicolich (1981, p. 790) postulated that symbolic play becomes independent of

play materials only when the player is able to both plan and mentally construct an action

before embarking on the activity. In addition, at this level, the ensuing play is

233


independent of context. Integration of a number of schemes (including rapidly

developing language) is necessary for successful symbolic activity at this sophisticated

level - a process described by McCune-Nicolich as "multischeme combinations" (1977,

p. 93).

7.15.1.2 'Thematic Content': Dimension Two

As with other symbolic-play behaviours, Dimension T w o ('thematic content')

follows a developmental sequence. Initially very young children engage in symbolic

play by representing themes in which they have regularly participated. At Level One

(according to Westby's Scale), when children are aged approximately 17 to 19 months

symbolic play reflects much of what involves them in their everyday life. It is common

to see toddlers of this age pretending to sleep or eat or wash - all of which are very

familiar themes to them and importantly, involve them. As with other behaviours,

thematic content in symbolic play progresses from that which is very concrete and

centres on the child, to that which represents activities children have observed

performed by the more familiar others in their environment. Finally in the later

preschool-years children are able to draw upon imagined or invented (and indeed often

unachievable in real life) themes for their symbolic activity.

First-born toddler participants exhibited significantly higher levels of symbolic

play in Dimension T w o ('thematic content') in Condition T w o (mixed-age play). There

were no other significant results for 'thematic content' in any condition.

It is useful to note, though, that Logistic Regression Analysis showed dyad play

and the unsatisfactory provision of play materials as the two explanatory variables for

the likelihood (the odds ratio) of participants engaging in more complex symbolic play

in Dimension Two, 'thematic content'. (The unsatisfactory provision of materials was

discussed earlier in 7.11.2 and the dyad condition in 7.7.1 and 7.8.)

T w o factors seem crucial when discussing participants' symbolic activity

reflecting thematic content (Dimension T w o ) behaviour. The first, as mentioned above,

is the child's stage of development and what this conveys in terms of expectations for

their expressions of thematic content. The second is the degree to which children's

development of 'thematic content' is supported by their experience in the play

environment - the toys, materials and equipment provided for their free-play.

234


7.15.1.3 'Organisation of Themes': Dimension Three

The participants exhibited higher levels of Dimension-Three behaviour

('organisation of themes') in Condition Three (dyad play); and secondly, when

complexity of symbolic play was examined, the younger participants (those aged from

17 to 19 months) were mostly likely (again in Condition Three) to exceed the expected

levels for their age, in all four dimensions of symbolic-play behaviours. Such results

are inconsistent with expectations according to Piagetian theory but receive support

from the research of others.

For example, Fenson and Ramsay (1980) investigated two dimensions of

symbolic play - organisation and decentration (in Westby's terms, 'organisation of

themes' and 'self-other relationships'). Their results showed that children were able to

perform more than single steps in the organisational dimension of symbolic activity

below the age at which sequential activity would be expected to occur. Fenson and

Ramsay postulated that this more complex behaviour might be the result of preplanning,

but could also be 'triggered' by the child's first action (1980, p. 177). For example if a

participant picked up a jug, then a pouring action, followed by a drinking action, may be

triggered. Nevertheless, as Fenson and Ramsay emphasise, for ordered actions to occur,

an ability to relate two or more actions is indicated (1980).

Mandler and McDonough (1995) in their study of infant long-term recall also

questioned the efficacy of 'triggering' in the very young child's recall process. They

postulated that, although observation alone can effect an association between an object

and sensori-motor action, practice is required for the infant to 'learn' that association.

Such evidence - namely, the ability to recall two or more actions combined with

experiential practice is required for successful and elevated activity in the symbolic play

dimension of 'organisation of themes' - supplied by Fenson and Ramsay (1980) and

Mandler and McDonough (1995) suggests that the 'triggering' theory in relation to

enhanced levels of 'organisation of themes' (the third dimension of symbolic-play

behaviour) can be discounted.

7.15.1.4 'Self-Other Relationships' (Decentration): Dimension Four

As discussed in Chapter Three (3.4.1), in amendment to Piaget's predictions,

McCune-Nicolich (1981) suggests that young children achieve decentration earlier than

18 months of age. Westby (1991), who appears to follow Piagetian theory, suggests

that children aged 17 to 19 months certainly engage in self-directed play. However, as

235


reflected in Westby's Scale for Assessing Children's Pretend Play (1991), the

expectation is that children have, by 19 to 20 months, progressed to including objects or

other people as passive recipients of their symbolic activity.

The current results showed that the youngest age group (17 to 19 months)

engaged in activity reflecting Dimension Four symbolic behaviour ('self-other

relationships') exceeding what was expected for their age. This occurred in both same-

age play (Condition One) and, more strongly, in Condition Three (dyad play).

As mentioned earlier, Fenson and Ramsay (1980) also investigated the

development of object-directed activity in young children's symbolic play. Their

findings showed that such activity emerged at the same time as self-directed activity,

and decentration was seen in children as young as 13 months. These results are in

keeping with the neo-Piagetian theorists' propositions. The current findings support

those of Fenson and Ramsay (1980) since our results indicate that participants exhibited

higher levels of Dimension Four ('self-other relationships') in Condition Three, dyad

play.

The above discussion reveals that explanations from Vygotskian and Piagetian

perspectives are complimentary. Furthermore, commonalities exist between the two

philosophies including a respect for individual differences and the belief that children

are active in their learning through participation and problem-solving.

7.15.2 Neo-Piagetian Perspectives

Consideration of neo-Piagetian perspectives such as those proposed by Halford

(1993) and Case (1991) provides explanations based on the information-processing

component of these perspectives. Each theorist perceives the child to be active in the

gathering and ordering of information aided by increases in basic capacities - for

example, working or short-term memory.

7.15.2.1 'Organisation of themes': Dimension Three from the Neo-Piagetian

Perspectives

Participants, in this current study, w h o exhibited a short but ordered sequence of

actions in Dimension Three ('organisation of themes') - symbolic behaviour thought to

be ahead of that expected for their age - may be experiencing what Case (1985)

describes as 'automisation'. Thus the acquisition of a new skill relinquishes short-term

memory capacity leaving it available for the acquisition of other, new abilities.

236


In a similar vein Halford (1993) suggests that young children often perform in

ways ahead of, not akin to, their understanding. Such performance may be related to

Halford's notions of 'knowing how' (a function of the short-term memory) as opposed

to 'knowing that' (a function of long-term memory). It is noted that in Mandler and

McDonough's study of infant's long-term recall 14-month-olds reproduced "two-action

novel causal events" (1995, p. 471) - through observational learning only at 11 months.

(Arbitrary events though were quite rapidly forgotten.) Additionally, Mandler and

McDonough's results support the notion that infants' long-term memories are able to

efficiently manage larger amounts of information than was earlier thought - for example

the 16 month-old infants in their research recalled "three-action causal events" (1995, p.

472). Further support is gathered from Hanna and Meltzoff's (1993) research in which

14-month-olds recalled demonstrated actions 48 hours later and importantly in a

different context from the one in which they observed the modelled behaviours. The

findings in the present study in conjunction with those of Mandler and McDonough

(1995) appear to support Halford's (1993) theory of long-term memory function - the

'knowing how' function.

Finally the current findings also seem to support the skill-theory framework

approach of Fischer and Farrar (1987). Enhanced symbolic-play skills were evident in

Condition Three (dyad play) a supportive and familiar environment. Fischer and Farrar

(1987) postulated that such an environment is conducive to an 'optimal level' of

performance as opposed to that which occurs in a 'sub-optimal' context.

In conclusion, while components of the information-processing approach take

precedence in explaining the results from a neo-Piagetian perspective, it is noteworthy

that both Case and Halford recognise the role of the socio-cultural context in children's

learning and development.

7.16 THE RESULTS FROM OTHER THEORETICAL PERSPECTIVES

In Chapter Three the perspectives of theorists Erikson, Bandura, Bruner and

Bronfenbrenner were explored as each offered information pertinent to the current

study. The findings which accord with several of the propositions seen as central to

their theories of Erikson, Bruner and Bronfenbrenner will now be discussed.

(Contributions arising from Bandura's theoretical position are presented in part 7.21.4

of this chapter when directions for future research are suggested.)

237


It is likely that benefits for children's psychosocial development arise out of

early acquisition of symbolic-play skills. Pretend or symbolic play - a strategy

recognised by Erikson (1963) as useful for the toddler engaged in resolving the

psychosocial crisis of 'autonomy versus shame and doubt' - provides an arena in which

the child can safely explore and practise the skills of autonomy devoid of shame and

doubt.

In the current project, first-born children showed an increase in the frequency of

symbolic play and displayed higher levels of symbolic play in Condition T w o (mixed-

age play) than later-born participants. These findings accorded with the literature

acknowledging the positive influence of first, parental-child attachment and secondly,

the efficacy of parental involvement in enhancing young children's play (Dunn & Dale,

1984; Slade, 1987a, 1987b; Meins et al., 1998). Such findings also concur with the

psychoanalytic perspective of Winnicott (1971) who, in the Freudian tradition, both

recognised the importance of the mother/care-giver child relationship and the benefits

thereof, and endorsed the notion that symbolic play emerged as infants experienced

'separateness' from their mothers or caregivers.

It is timely to n o w to turn to some ethological aspects of the study - those that

correspond with Bruner's notions of both play and its development. Toddler-aged

participants displayed more frequent and higher levels of symbolic play when engaged

in free play with older peers than in free play with their same-age peers. These results

support Bruner's (1972, p. 697) concept of 'tutor-proneness' - whereby children are

thought to be especially sensitive to the input of others' in their immediate environment.

In dyad play the older peer's communicative behaviour (though lacking

statistical significance) appeared instrumental in supporting the toddler's activity.

Importantly this communication appeared to be a stimulus, (and at times provide the

necessary structure - in Bruner's terms 'scaffold') for the occurrence of 'joint action'

and 'joint reference' described by Bruner as essential for the development of language

and social skills (1975).

As stated earlier in Chapter Three (3.7) from Bronfenbrenner's (1979)

ecological theoretical stance the L D C C C comprises a component of the participant's

'microsystem'. Secondly, the L D C C C is a context ideally suited to providing

opportunities for the occurrence of "proximal processes" thought by Bronfenbrenner to

be critical to the child's well being (1993, p. 8). In addition he considers that

predictability and stability in everyday life are desirable for young children

238


(Bronfenbrenner, 1993). The present results confirm that younger participants benefit

from engagement in both mixed-age play and dyad play - each condition likely to

support the occurrence of "proximal processes" (1993, p. 8). Finally perhaps the

benefits of stability that Bronfenbrenner proposes as essential for young children, are

illustrated in the current results showing enhanced symbolic-play activity by those

participants attending L D C C C s for four or more days a week.

7.17 VARIABLES LIKELY TO HAVE INFLUENCED THE RESULTS

7.17.1 The Provision of Play Materials

In Chapter Six (6.11.2.1) the provision of toys, materials and equipment, in the

L D C C C s to support the development of symbolic play was assessed and this provision

summarised in Table 6.16.

It was anticipated that a relationship would exist between the satisifactory

provision of play materials and both the enhanced levels of symbolic play and the

increased frequency of symbolic-play activity. This prediction was, however, not

supported since the unsatisfactory provision of play materials was found to influence

participants' complexity of play. (This finding was discussed earlier in this chapter in

7.11.2.)

Nevertheless, the research of others (Westby, 1991; Howes & Unger, 1992;

Howe, Moller, & Chambers, 1994; Rusher, Cross, & Ware, 1995) indicates that the

provision of appropriate toys, materials and equipment is likely to support the

development of symbolic-play. (Possible reasons for this discrepancy were explored

earlier in part 7.11.2.)

Each of the significant results, higher levels of the two symbolic-play behaviours

'organisation of themes' and 'self-other relationships,' were outcomes of the

participants' play in Condition Three (dyad play). Significantly more events (an

increased frequency) of symbolic-play activity also resulted from the toddlers'

participation in Condition Three.

Three standard sets of toys were provided for play in Condition Three. The

toddler participants were observed in Condition Three (dyad play) on three separate

occasions. Different materials were provided on each of these instances. A detailed

description of these toys together with the type of symbolic play likely to be elicited by

their use was presented in Chapter Five (5.6.2.7).

239


In summary and in accordance with the literature (Westby, 1991; H o w e et al.,

1994; Rusher et al., 1995) two ranges of play materials were presented. T w o examples

of the first range of play materials, familiar and highly realistic and likely to support

emerging representational skills, were presented to the dyads for the first and third

session of dyad play (Condition Three). The first set consisted of two dolls, a plastic

tea-set with cutlery and a coffee pot; the second set contained play-dough with

appropriate accessories (including bowls, spoons and measuring cups). The second

range of materials likely to support "role and situational transformations" (Westby,

1991, p. 147) comprised table blocks, wheeled vehicles and human-like figurines.

These were presented for the second session of dyad play in Condition Three.

The materials presented for the first and third session of dyad play, the tea-set

and accessories and play-dough and associated equipment, are likely to have supported

the significant results in Dimension Three ('organisation of content'). This dimension

of symbolic-play behaviour refers to the sequencing of logically connected actions

engaged in by a player. In the early stages of symbolic play children often enact one

action only, followed by another unrelated (random) action. As symbolic-play activity

increases in complexity, children engage in a number (maybe only two or three initially)

of sequenced or connected actions - for example, stirring play-dough, spooning it out

(or moulding it by hand) into smaller containers (ice-cube trays were supplied in this

study) and then perhaps placing it in an oven. Play with play-dough or the tea-set and

accessories supports the occurrence of 'organisation of themes'. At this point the writer

is reminded of the Vygotskian idea of 'societal and cultural rules within play', Fenson

and Ramsay's (1980) notion of 'triggering' and the neo-Piagetian premise that the

capacity of the working memory supports the process of 'automisation' (Case, 1991).

Each of those is an explanation for an increase in 'organisation of themes'. However,

central to each of those ideas is development in terms of the child's intellectual ability -

such actions cannot occur unless the child is cognitively relating two or more actions.

A significant relationship between Dimension Four ('self-other relationships') of

symbolic play and Condition Three (dyad play) was found. It is likely that the materials

presented in the first and second sessions of dyad play, dolls in the first session and

human-like figurines in the second, lent substantial support to the players' symbolic

activity in Dimension Three.

Children's first manifestations of the symbolic behaviour 'self-other

relationships' clearly demonstrate their o w n involvement or participation. They are the

240


agents, the 'doers' in that they pretend to eat, drink, sleep and so on. Having completed

this stage of symbolic activity, children then progress to acting on 'others' - objects first

and later, other humans. However the provision of 'others' (such as dolls or human-like

figurines) provides children with the option and opportunity to act upon others. These

possibilities were available in Condition Three in the first two sessions. In hindsight the

materials supplied for session three in dyad play were perhaps not sufficiently

conducive to promoting enhanced levels of 'self-other relationships' symbolic-play

behaviour, as there were no props likely to suggest to or tempt the toddler to act upon

'others'.

The experience children bring with them to a situation is also of interest. The

toys for same-age play (Condition One) and mixed-age play (Condition Two) were

those provided by the L D C C C staff to support the program they had planned for the

children. A n assessment of those together with a summary (Table 6.16) was presented

in Chapter Six (6.11.2.1).

7.17.2 Features of One Participating LDCCC

One participating L D C C C , Centre Nine, it will be recalled failed to gain

accreditation. The provision of play materials to support symbolic play was assessed (in

accordance with the guidelines suggested by Westby, 1991; Howes & Unger, 1992;

Rusher et al., 1995) as unsatisfactory. Five participants attended this centre which

achieved the highest rating awarded on the Environment Rating Scale. The participants

(two of w h o m were in the youngest age group - 17 to 19 months) exceeded their

'expected' level of symbolic play on more than 50 percent of the observed occasions.

Interestingly these events were distributed almost equally across the three conditions

(same-age play, mixed-age play and dyad play).

The play environment at Centre Nine appeared to lack the abundance of play

materials generally on display in L D C C C s . However, household items and objects

donated by parents and others were plentiful. Consequently it was often very possible

for two or three children to pretend to make coffee (each using real coffee plungers) or

to 'cook' using real frying pans! Westby (1991) and others (Howe et al., 1994; Rusher

et al., 1995) endorse the provision of life-like or high-realism props to effectively

support emerging symbolic play.

Furthermore, staff-child ratios in Centre Nine were higher than those required by

N e w South Wales State regulations. In same-age play (Condition One) the average

241


ratio was one adult to 3.5 children (1.5 children less than required by regulation). In

mixed-age play (Condition T w o ) the average was 5 children to one adult - again fewer

children per adult than required by law.

T w o components of the daily program were noticeable as supportive of young

children's play. First the day was planned in such a way as to allow for quite extensive

periods of play. Same-age free play and mixed-age free play periods were each

allocated one hour and forty-five minutes. Though, other events such as planned small-

group activities and routine activities such as morning-tea and nappy-changing took

place within these extended free-play sessions. Research by Tegano and Burdette

(1991) and Christie and Wardle (1992) indicates the need for extensive periods of time

to be allocated for children's play. Shorter sessions of 30 or 40 minutes are inadequate

- perhaps more especially so with young toddler-aged children w h o often spend

considerable amounts of time exploring the provisions for play before actually

becoming engaged in playful activity.

Secondly, in Centre Nine as a result of the withdrawal of government subsidies,

children's attendance patterns were reduced but not sufficiently to affect staffing levels.

However, the reduction in numbers of children attending allowed flexibility. For

example siblings were often together at a number of times during the day as opposed to

being separated except for specifically the mixed-age free-play periods. A similar

leniency occurred for a number of older preschool children who spent time (almost at

will) in either the 'toddler' room or the 'baby' room.

Staff profiles at this centre were interesting. Apart from those trained in Early-

Childhood Education there were four other staff with tertiary qualifications including

accountancy, social welfare and the creative arts. (Only one member of staff was in the

'untrained' category.) While one member of the staff failed to complete the staff

questionnaire, the remaining nine who completed the questionnaire returned a 90

percent positive response rate to the four questions relating to children's play. The staff

at this centre were observed to be actively involved with the children in their play and

activities and rarely seen in a purely supervisory role. In fact Centre Nine was awarded

three years accreditation shortly after the data collection was completed.

A second explanation may be in keeping with the research of Steel and Hrncir

published in 1985. A s mentioned earlier these researchers were interested in the degree

to which high-realism (termed by them as 'high prototypicality') play materials and

adult play suggestions elicited symbolic play by 24-month-olds. These researchers

242


found that the adult's use of a 'low-prototypical' prop in play appeared to enable the

toddler to subsequently use that prop in a representational manner (1985). It is possible,

given the conditions prevailing in Centre Nine, that the participants were 'enabled' in

transformation (or representational) ability by adults modelling the use of low-realism

props.

The participants' positions in their families m a y also offer some insight to the

current dialogue. Three of the participants at Centre Nine were first-borns. Lyytinen,

Poikkeus, and Laakso (1997) found first-borns to exhibit significantly more symbolic-

play behaviour at 20 months of age than those not first-borns. It m a y be that the three

participants under discussion here have more opportunity for play partnerships with

their parents and others in their environment. Such experience m a y contribute to their

symbolic-play skills supporting their engagement in enhanced levels of symbolic play

with their older peers in dyad play.

Finally, as discussed earlier in 7.11.1, the number of days participants attended

L D C C C was found statistically to be associated with the 'likelihood' of their exhibiting

more complex symbolic play than would be expected for their age. Four of the five

participants attended Centre Nine for at least four days a week - the fifth participant

attended for a minimum of three days a week but attended 'casually' if vacancies

allowed on the remaining two days.

7.18 RESULTS INCONSISTENT WITH THE PREDICTIONS

7.18.1 Unsupported Predictions

As stated in the major hypothesis the expectations were that toddlers w h o were

engaged in free play (either mixed-age play or dyad play) with older preschool-aged

peers would exhibit symbolic play first more frequently and secondly, at higher levels

than those toddlers not engaging in free play with older preschool-aged peers. The

results indicate support in part only for the researcher's predictions. The following

predictions remain unsupported:

- That participants would exhibit symbolic play more frequently in

Condition T w o (mixed-age play);

- That participants would exhibit higher levels in three dimensions of

symbolic play ('thematic content', 'organisation of themes' and 'self-

other relationships') in both Condition T w o (mixed-age play) and

243


Condition Three (dyad play);

- That first-born participants would exhibit events of symbolic play more

frequently than 'later-born' toddlers in Condition Three (dyad play);

That first-born participants would exhibit higher levels of symbolic

play than 'later-born' toddlers in Condition Three (dyad play).

7.18.2 Results Not Consistent with Predictions

The following results are inconsistent with the predictions:

- There were no significant differences between the means for the

frequency of symbolic-play events in Condition T w o (mixed-age play);

- There were no significant differences between the means of the levels

of Dimension One ('decontextualisation') Dimension T w o ('thematic

content') in any of the three conditions (same-age play, mixed-age play

and dyad play);*1

- There were no significant differences between the means of any

dimension of symbolic play in Condition T w o (mixed-age play); *2

- Contrary to the researcher's expectations an unsatisfactory provision of

play materials was found to be an explanatory variable in the likelihood

(the odds ratio) for a participant to show more complex symbolic-play

behaviour. This finding was explored earlier in this chapter in 7.11.2.

(*1 When the relationship between birth order and levels of symbolic play was

tested, significant results were shown in mixed-age play [Condition Two] in all four

dimensions for first-born participants.)

(*2 W h e n the relationship between age group and levels of symbolic play was

tested, significant results were shown for participants in the 20-22 month age group in

Condition O n e [same-age play] in two dimensions.)

244


7.19 DISCUSSION: THE CONTEXT OF THE CURRENT ENQUIRY

The possible reasons why the predictions indicated above were not supported by

the results of the study are discussed below in the context of this study.

7.19.1 The Play Environment

7.19.1.1 Play Materials for Conditions O n e and T w o (Same-Age Play and Mixed-

Age Play)

As discussed in Chapter Six (6.11.2.1) and summarised in Table 6.17, the toys,

materials and equipment supplied by the L D C C C s for Conditions One and T w o (same-

age play and mixed-age play) were assessed - in accordance with Westby's guidelines -

in terms of their supporting the development of symbolic play. (Of the eight

participating centres, six were assessed as satisfactory and two as unsatisfactory.)

Westby (1991) considers the pairing of construction and small representational

materials as most beneficial in supporting young children's pretend play. In the present

study, such provision was seen for more than 50 percent of the observation sessions in

five centres only. One other centre provided representational materials irregularly,

while provisions for this type of play were never seen at two other participating centres.

Thematic play materials, other than those supporting domestic or home-related

activities, were noticeably absent in the outdoor environments of five participating

L D C C C s . T w o centres regularly had areas equipped to encourage the development of

the thematic dimension of symbolic play, while one centre catered for thematic play on

an irregular basis.

Five centres failed to set up provisions for such play on any observed occasion.

T w o of these five centres achieved a rating of 'unsatisfactory' on the Environment

Rating Scale while a third was rated as 'adequate' only. It may be that the lack of

suitable and appropriate toys, materials and equipment in conjunction with the low

ratings on the Environment Rating Scale impacts negatively upon the children's

experience.

In conclusion the results lacking significance recorded for Dimension T w o of

symbolic play (thematic content) perhaps indicate that the L D C C C environment may

not have been conducive to the children's (toddlers and older peers) taking advantage of

potentially supportive play materials. Implications arising from this are presented in

part 7.21 of this current chapter.

245


7.19.1.2 Play Materials for Condition Three (Dyad Play)

As mentioned earlier, guidance from the research of others (Westby, 1991;

Howes & Unger, 1992; H o w e et al., 1994; Rusher et al., 1995) was heeded when toys,

materials and provisions were chosen for the three dyad-play sessions. T o that end,

props supporting thematic play were provided for two of the dyad-play sessions

(Condition Three).

A 'tea-party setting' was set up for the first session. These materials were

chosen as likely to represent experiences familiar to either the participants or others

familiar to them - for example tea-time or coffee-time at home. Only one dyad-couple

used these materials as props for 'restaurant' play - a more complex theme than that of

tea or coffee-time at home.

Table blocks and wheeled vehicles, which according to Westby's (1991)

recommendations were likely to elicit more complex thematic play, were provided for

the second dyad session.

O n reflection, these first two sets of materials presented to the children in

Condition Three (dyad play) - though effective in promoting thematic symbolic play -

may have prompted higher levels of the symbolic-play behaviour of 'self-other

relationships' (Dimension Four) as children had access to 'others' - namely, dolls and

human-like figurines. N o similar props were supplied in the third 'play-dough' session.

Consequently participants lacked objects upon which to act. While the preschool

partner was available as a potential recipient for the toddler to 'act upon', the context

and physical setting in which the dyad-play situation occurred may not have been

conducive to such activity.

Again in retrospect, it may be that the tea-set play and play-dough activities may

have been more closely linked with the child's own experiences than the activity

involving table blocks, wheeled-vehicles and figurines. It is possible that those

materials (for tea-party and play-dough play) were closely related to familiar activities

in the everyday lives of many of the participants - both activities represented common

domestic themes. This factor may have affected the toddler participants' ability to

move beyond auto-symbolic activity (Dimension-Four behaviour) to acting upon

'others' - either objects or people.

246


7.19.1.3 The Context of Mixed-Age Play (Condition T w o )

The prediction was that participants would exhibit both higher levels and more

events of symbolic play in Condition T w o (mixed-age play). These expectations were

not supported by the results with the exception of those highest levels of symbolic play

exhibited by first-born participants.

A n unfounded concern when developing the methodology for the current

enquiry was that in mixed-age play (Condition T w o ) the younger and older participants

might not play freely together - that some degree of age segregation may exist.

Other explanatory factors in the child-care day that were noted include: the

social context of mixed-age play; the influence of the adults' interaction and teaching

styles; group numbers and staff-child ratios; and access to and availability of play

materials. While these factors are each clearly important influences on children's

symbolic play, they were either beyond the scope of this study, in terms of collecting,

measuring and statistically analysing data, or not possible to control. However, they are

worthy of the descriptive discussion that follows.

The social context of mixed-age play varied considerably and was dependent

upon a number of features, which influenced the children's experience during that

period. First, sibling 'couples' noticeably spent considerable periods of time together.

(These relationships were observed and recognised during the process of identifying the

participating toddlers' older familiar peer for dyad play). Secondly, those toddlers who

had a 'chosen older familiar peer' also spent considerable amounts of time in his or her

company. However, toddlers who were not in 'partnership' with an older child could

often experience what appeared to be quite an anxious time spent either avoiding the

older more boisterous children or engaging in repeated and unsuccessful attempts to

gain a 'turn' on equipment or acquire toys with which to play.

7.19.1.4 Group Numbers and Staff-Child Ratios

A N e w South Wales Regulation stipulates the following ratios in homogeneous

groups: 0-2 years - one adult to five children; 2-3 years - one adult to eight children;

and 3 to 5 years - one adult to ten children.

In this study when the toddler participants (aged 17 to 31 months) were in

mixed-age play (Condition T w o ) their ages ranged from the youngest toddler to

(depending on the time of year) children w h o were close to six years old. Mixed-age

free-play sessions were generally characterised by greater numbers of children present

247


compared to those in attendance in the same-age free-play sessions (Condition One).

Large numbers of children may themselves be quite overwhelming and even

intimidating for the younger ones in the group. (The most children observed in a

mixed-age play period was 36, compared to an average in same-age free play of 11.)

In mixed-age play, while maintaining the correct ratios within age groups, staff

are able to cater for a larger number of children than would otherwise be possible. For

example 2-year-olds in the youngest group - birth to 24 months - require one adult to

every group of five children. However, if the 2-year-olds are being catered for as part

of the 2- to 3-year-old group, then the one adult to eight children ratio is permissable.

In all but two of the participating L D C C C s , in the outdoor mixed-age play sessions the

greater part of staff activity appeared to be monitoring 'turn-taking' on equipment or

settling disputes. Consequently children's access to staff for either assistance or playful

interaction was reduced and overall opportunities for staff contact were limited.

Contexts where the adult-to-child ratios were low were found to be more

effective in eliciting very young children's imitation of linguistic and gestural

behaviours modelled by adults (Francis & Self, 1982). Results such as these emphasise

the need to make available more-skilled or more-accomplished 'others' such as older

peers accessible as models and demonstrators of a range of behaviours appropriate for

the younger children's acquisition.

7.19.1.5 Access to Materials

A s reported in Chapter Six (6.11.2.1) the provision of materials to support

symbolic play was unsatisfactory in two of the participating L D C C C s and satisfactory

in the remaining six centres. In mixed-age play (Condition T w o ) the range of materials

provided for play was likely to be developmentally appropriate - catering for the needs

and interests of all groups - but the number of toys, materials and equipment available

for each child's use was often reduced and the competition for use greatly increased.

Hence opportunities for the younger participants to engage with materials supporting

symbolic play were often limited.

248


7.20 IMPLICATIONS FOR EARLY-CHILDHOOD PRACTICE ARISING

FROM THE CURRENT STUDY

A number of implications for early-childhood practice arise from the results of

the current project. The ensuing discussion explores the results from both contextual

and theoretical perspectives. The major implication is the usefulness of dyad play as a

social context for young children's learning and acquisition of skills. It appears likely

that the dyad context offers increased efficacy if the older play partner is familiar to, and

where possible 'chosen' by, the younger peer.

While to some extent implications are contingent on adequate staffing,

opportunities for mixed-age play and dyad play can be encouraged and supported in a

number of ways. First, the thorough and careful observation of children in a free-play

context informs staff of children's relationships and interests. Subsequent thoughtful

planning of the environment and the components within that environment can then

provide opportunities for relationships between older and younger peers to strengthen.

Implications arise too for the provision of time, space and materials for play and

adult involvement in play.

7.20.1 Mixed-Age Free Play

First the results of this current examination of influences on toddlers' symbolic

play supports the practice of younger children engaging with older peers in mixed-age

free play - whether in a group or dyad situation. The older peer in dyad play is able to

engage in tutoring and/or 'scaffolding' activities that enable the younger children to

achieve more than they would independently. The context of dyad free play is also

favourable to social interactions comprising elements of 'joint attention' and 'joint

activity' identified by theorists and researchers as so essential for the child's acquisition

of new skills and learning. Dyadic situations appear to be particularly effective for the

younger child to engage in imitation of play behaviours modelled or demonstrated by

the older partner.

Partnerships between toddlers and older peers (either as dyads or as partners in

small mixed-age groups) can be fostered in L D C C C s , where staff closely observe and

plan for individual children's development, as part of the regular daily program. (Staff

are also well-placed for identifying younger peers' favoured or preferred older play

partners.) The practice of L D C C C staff and teachers is, when planning for young

children, to first, identify their developmental characteristics and secondly, their

249


interests. In doing so a developmentally appropriate task identified along the child's

developmental 'pathway' and in keeping with each child's interests constitutes a goal

within his or her 'zone of proximal development'. Such tasks reflecting symbolic-play

development can be readily recognised by personnel skilled and practised in child-

development theory and research as applied to play.

In addition, to ensure successful dyad partnerships a number of environmental

components, including time, space and materials, require the attention of the L D C C C

staff

7.20.2 Time for Play

It is essential that adequate time for play to develop be allocated. Christie and

Wardle (1992) suggest a minimum of 30 minutes for free play in preschool and

kindergarten rooms - longer periods may be warranted for younger children who often

thoroughly explore play materials before engaging in playful activity. In the current

study, free-play periods of almost two hours in one L D C C C were thought likely to

contribute to the significant results achieved by participants. Limited time for play can

result in children choosing not to engage in activities that require exploration, or

preparation and planning, before play is effected.

7.20.3 Space for Play

Successful play can also be contingent on the area or space used for play

activities. In Australia the outdoor environment facilitates freedom of expression and

activity and is, when equipped with a range of toys, materials and equipment to support

the development and complexity of children's symbolic play, an appropriate arena for

such play (see Appendix Fii).

However, it is thought that a beneficial component of the current enquiry was

the setting up of the dyad condition a little aside from the children's main activity area.

This can be accomplished either indoors or outdoors but, as with all activity areas, close

supervision and monitoring by staff is essential.

7.20.4 Toys, Materials and Equipment for Play

Materials to support the development of symbolic play need to be carefully

chosen with an understanding of the stage of symbolic activity children are currently

experiencing. For example, careful and informed judgement is necessary when

250


assessing the degree of realism of props required. A n explanatory variable for the

enhanced complexity of participants' symbolic play in the current project was an

unsatisfactory provision of play materials. It is useful to assess (though observation of

children's play) the degree of realism required and whether or not it is time to reduce

very life-like materials and replace them with objects that lend themselves to

transformation (e.g. providing ice-cube trays for possible use as patty cake pans). It

may be important to recognise the time to limit the number of life-like props while

ensuring materials suggesting representation are provided.

Guidance can be taken from research suggesting that the level of structure in the

materials and activities provided for children can impede their play activity (Tegano et

al., 1991) while adult modelling with materials lacking life-like qualities can assist

children's transformation of objects and representational play (Steele & Hrncr, 1985).

7.20.5 Adult Involvement in Symbolic Play

A relationship between first-borns and the frequency of symbolic-play events

and the levels of symbolic play was found in the current study. First-born participants

exhibited more frequent and higher level symbolic play in mixed-age play suggesting

they were perhaps 'more prepared' for, and more experienced in symbolic play than

later-born participants. As the literature reviewed in Chapter T w o (2.6) informs,

maternal activity and interaction contributes much to very young children's symbolic

play activity. Part of the first-borns' success would appear to be linked to the

attachment between mother and child.

While L D C C C s constitute a very different context to that of the home or even

that shared by a mother and child in a play laboratory there are some aspects of the

mother's behaviour that could perhaps be adopted by caregivers in L D C C C s . For

example, working with mothers and their 31-month-olds w h o were securely attached,

Meins et al. (1998) found the children were able to successfully integrate play

suggestions into their symbolic play. "Mind-mindedness" (Meins et al., 1998, p. 20) - a

way of responding to children as individuals with minds - and a maternal style of

interaction were identified by the researchers as likely to be influential.

Enz and Christie identified three types of 'teacher play interaction styles' -

"stage manager, co-player and play leader" (1997, p. 71) - as being most supportive of

children's play. (Perhaps future research will investigate whether or not these play

interaction styles coincide with the play techniques adopted by parents.) It would be

251


useful for L D C C C staff to be trained in these play interaction styles in preference to

others ("interviewer/narrator and director", 1997, p. 62) found by the researchers to

possibly impede young children's play.

Finally it is proposed that adult involvement could be effective in guiding and

extending young children's symbolic-play activity. Participation in children's play as

suggested by Enz and Christie (1997) would provide opportunities for staff to scaffold

(both verbally and through actions) their symbolic play.

In order to promote children's symbolic activity it is useful to consider activities

and techniques that are effective in extending or prompting activity in each dimension

of symbolic play ('decontextualisation', 'thematic content', 'organisation of themes'

and 'self-other relationships') as identified by Westby (1991).

W h e n adults are engaged in symbolic play with young children, modelling or

demonstrating the following behaviours assist children to extend their symbolic-play

skills: encouraging the substitution of one object for another ('decontextualisation');

introducing themes and scripts in which familiar others engage ('thematic content');

sequencing or linking actions and events ('organisation of themes'); and finally acting

upon 'others', either objects (dolls) or people ('self-other relationships').

Steele and Hrncir's (1985) work implies that children are very receptive to

adult's pretend-play ideas and providing these are developmentally appropriate and in

Vygotskian (1987) terms are just a little ahead of the child's development then

monitoring children's play and, when and if appropriate, unobtrusively commenting or

intervening or making suggestions could support the development of children's

symbolic activity.

7.21 DIRECTIONS FOR FUTURE RESEARCH

7.21.1 The Influence of Adult Interaction and Teaching Styles

The L D C C C social environment contains a number of adults all of w h o m bring

with them a range of different attributes including: interaction and teaching styles; ages

and socio-cultural backgrounds; training and qualifications; and personal qualities,

strengths and limitations. The systematic identification of the influence of such factors

relating to the care-giving personnel on toddlers' symbolic play was beyond the scope

of this present enquiry. However, the literature reviewed in Chapter T w o (2.9) suggests

that the interaction and teaching styles of adults are first, related to training and

252


qualifications in Early-Childhood Education and secondly, are such that may comprise a

substantive influence on young children's cognitive development - for example,

symbolic play skill acquisition. Clearly consideration of interaction and teaching styles

would be a useful component of any future research in a similar vein to this present

enquiry.

Such a recommendation receives support from the following research outcomes

including: training in child development and communications affected both care-giver

attitudes and their behaviour towards young children (Arnett, 1989); children (in this

case with special needs) engaged in more significantly developmental behaviours when

the adults in their learning environment were "responsive, supportive and relatively

nondirective" (Mahoney & Wheeden, 1999, p. 67); teacher play interaction styles

indicated that trained and experienced teachers chose strategies that were "effective in

promoting the children's make-believe play" (Enz & Christie, 1997, p. 62); untrained

and/or inexperienced staff adopted strategies that did not support (and at times impeded)

children's play (Enz & Christie, 1997); and early-childhood training accounted for 62

percent of the variance in teacher inputs relating to excellence and high provision of

care (Honig & Hiralli, 1998).

Additional literature considering the strength of the emotional relationship

between care-giver and child clearly indicates that benefits accrue to the child (Howes,

Hamilton, & Matheson, 1994; Howes & Smith, 1995).

7.21.2 The Influence of the LDCCC Environment on Play and Peer Interactions

The quality of care-giving in the participating L D C C C s was assessed (as

unsatisfactory, adequate or good) in the current investigations but is worthy of closer

examination in future research. The quality of the L D C C C care-giving environment

impacts on both children's play and their social interactions, the latter being an essential

component in children's successful development of play.

As described in Chapter Two, 2.10, relationships exist between families and

L D C C C influences. Preschool-aged children in high-quality care-giving environments

were shown to be more social and effective at problem-solving (Holloway & Reichhart-

Erickson, 1988) and both peer-play development and the complexity of that play were

found to be enhanced in L D C C C providing high-quality care (Aureli & Procacci, 1992;

Howes & Matheson, 1992; Aureli & Colecchia, 1996).

253


Lower staff-to-child ratios are usually a feature of L D C C C offering quality care

giving. Substantial differences in the way that children used their time related to the

centres' quality of care-giving was shown in the research by Vandell and Powers (1983)

and Howes, Phillips, and Whitebook (1992).

7.21.3 The Influence of Older-Peer Communicative Behaviour

Observations during the trial period confirmed the expectation that the older

peer would be likely, during dyad play, to offer the younger child information and

support relating to any ensuing play. While beyond the scope of this study it seemed

useful to note the older peer's communicative behaviour in accordance with a scale

modified from that developed by Beizer and Howes (1992) - see Appendix Giv. These

researchers considered a range of maternal behaviours including verbalising, physical

assistance, elaboration, encouragement, instruction and clarification as 'indirect'

support. Direct instruction or invitations, gestures and a combination of both of these

behaviours were classed as 'direct' support.

The older peer's communicative behaviour was noted for every 30-second

period during Conditions T w o and Three. If both 'indirect' and 'direct' behaviours

were evident in a 30-second period then the 'direct' behaviour (as the highest, most

informative behaviour) was recorded. Categories were pooled so older-peer support

behaviours were classified as either 'indirect' or 'direct'. The Table contained in

Appendix Gvi illustrates the frequency and type of communicative behaviours engaged

in by preschool-aged peers, when toddlers were displaying symbolic-play behaviours,

during Conditions T w o and Three (mixed-age play and dyad play). In summary there

was a total of 21 instances of 'indirect' and 175 instances of 'direct' communicative

behaviours in mixed-age play; and, a total of 63 'indirect' and 716 instances of 'direct'

communicative behaviours in dyad play.

Examples of direct pretend suggestions noted by the researcher include, "Here's

the sugar - take a big scoop" and from the same older peer, "I'll get some peas, and

some bacon now". These suggestions arose in the dyad condition when dough was the

play medium. A n example of indirect support from dyad block play was "This is where

it goes R . The car goes there." In the tea-party dyad-play setting the older peers'

questioning of the younger peers and their requirement for tea occurred in a number of

ways including, "Would you like some more tea?" and " N o w then, how about a cup of

tea." A more complex direct pretend suggestion, "Pretend it was a girl. Pretend they

254


fell out of the truck. N o w this one's pretending to be a girl" was observed in dyad play

when blocks were provided with small-wheeled vehicles and two human-like figurines.

Additionally, a familiar self-chosen older peer in dyad play is likely to provide

the younger participants with support that is relevant to their socio-cultural needs. T w o

studies (Tomasello, Savage-Rumbaugh, & Kruger, 1993) and (Carpenter, Tomasello, &

Savage-Rumbaugh, 1995), offer support for this view. Each of these research teams

found that the socio-cultural environment experienced by both enculturated

chimpanzees and children appeared to be an essential component for both groups to

succeed on the imitation tasks.

The communicative support offered by the preschool-aged partner(s) in mixed-

age play and dyad play could be an influential or explanatory variable for participants'

enhanced symbolic-play activity and warrants further investigation.

7.21.4 The Toddler Participants' Imitative Behaviour

The toddler participants' imitation of the older peer in both mixed-age play and

dyad play was also recorded whether or not the younger child was engaged in symbolic

play. Initially imitation was noted as either verbal or gestural (action) but as events of

imitation were limited these two categories were pooled. In mixed-age play when the

toddler was not engaged in symbolic play there were very few (five) events of imitative

behaviour. Similarly in dyad play when the participant was not engaged in symbolic

play, there was only one event of imitative behaviour. In mixed-age play when the

toddler was exhibiting symbolic-play activity more imitative behaviour was observed -

27 events in all. However, in dyad play when the participants were engaged in

symbolic-play activity there was a substantial amount of imitative behaviour - 116

events. Events of imitation are presented in Appendix Gvi.

It is interesting (and in keeping with expectations for children of this age and

stage of development) to note that that imitative actions were far more in evidence than

verbal imitations - twice as frequently observed in mixed-age play and almost three

times as frequently in dyad play.

It is tempting to speculate that imitation may have a much more important role

in the development of symbolic play than is as yet recognised. It appears to the

observer that often a young toddler engages in play with a life-like prop representing

actions in a way that is more akin to ritual than symbolic play. These ritualistic

behaviours, such as those observed when a toddler handles and 'uses' a replica mobile

255


phone or 'walks' a human-like figurine, exist before an observable pattern or episode of

symbolic play emerges. Piaget suggests that "ritualisation is preparation for for

symbolic games", (1962, p. 93) and occurs before deferred imitation is achieved (1962,

p. 95).

As described earlier in Chapter Three, Bandura (1977), a proponent of

observational learning, suggested that in order to acquire learning and skills through

observation children engage in the four sequential processes of attention, retention,

motor reproduction, reinforcement and motivation. Dyad play provides a medium

conducive to these processes.

First, toddler participants were likely to attend in the dyad condition as their

partners were 'self-chosen' familiar peers - this factor also supported a second

component in the observational learning process, motivation. The toddlers' attention

was maintained as dyad play was positioned a little apart from the children's main play-

activity area. Importantly, the younger participants' stage of development allowed them

to first, reproduce motorically any action modelled or demonstrated by the older peer

and secondly, retain the new information. Finally the older peers in the dyad

partnership often reinforced (either verbally or non-verbally) the toddlers for any

behaviours exhibited whether or not they included imitation.

In conclusion Condition Three (dyad play) elicited immediate imitative

behaviour.

7.21.5 Deferred Imitation

Identification or assessment of the toddlers' acts of deferred imitation was not

one of the goals of the current study. (Such an investigation would be useful and is

certainly one of a number of future directions arising from this research.) However, it

seems likely that, in keeping with the research (Tomasello et al., 1993; Hanna &

Meltzoff, 1993; Carpenter et al., 1995;), deferred imitation could have accounted for

some of the younger peers' play behaviour in the dyad-play session in the present study.

(Each of the studies referred to above is reviewed in more detail in Chapter 2, 2.4.)

Toys and play materials presented for the children's use in the dyadic condition

were unavailable in Conditions One and T w o (same-age play and mixed-age play). The

question arises whether or not toddler participants who engaged in higher levels of and

more complex, symbolic play with an older peer in dyad play (Condition Three) would

256


reproduce such play when engaged with others in either same-age play or mixed-age

play.

In the results, first-born toddler participants engaged in more frequent and higher

levels of symbolic play in mixed-age play (Condition T w o ) than in either same-age play

or dyad play - Conditions One and Three. It is suggested that these findings may be an

outcome of deferred imitation.

7.21.6 Benefits for Older Peers

The context of mixed-age play whether in a group or a dyadic situation

(comprising Conditions T w o and Three in the present enquiry) offers a number of

benefits for the older preschool-aged peer ranging from enhancing socio-emotional

development to the practise of language skills and communicative behaviours. Such

benefits are described in more detail below.

7.21.6.1 Benefits for the Older Peers' Socio-Emotional Development

Peller (1978) considered the benefits of mixed-age interaction from the older

child's perspective to be as important as those that accrue for the younger peer. She

recognised that the older peer, in company with a younger child, experienced a number

of roles including "the role of the older, the protecting friend, and at times the role of

the follower, the receiving party" (1978, p. 87). Peller proposed that such experience

"broadened the development of the child's ego and strengthens his feelings of

competence" (1978, p. 87). This notion is strengthened if children's psychosocial

development - according to Erikson's (1963) theories - is considered.

It is likely that play activity with younger, less-experienced peers may support

the older child in the resolution of what Erikson viewed as the third psychosocial stage

of 'initiative versus guilt' when children were aged from three to six years - the period

in which young children experience a sense of ambition and responsibility as they use

initiative. Experimentation occurs when, especially through play, children explore the

roles and attributes of the sort of people they wish to be (Berk, 1997). While

traditionally parents and family are the key social agents during this period, it may be

that for children attending L D C C C s the staff and peers encountered in the L D C C

environment m a y partially fulfil that role.

In the dyad-play condition older peers had opportunities to engage in leadership,

practise self-regulation and take responsibility for directing and managing the play

257


activity. Examples illustrating such older peer behaviours are taken from the

researcher's field notes and include, "You need to make it bigger," "This can be mine

and look this is yours," "This is going to be your ice-cream."

Importantly, the dyad situation allowed preschoolers to explore a range of roles

including the role described by Katz, Evangelou and Hartman (1990) of the 'expert' in

the company of a 'novice' (the toddler). In Vygotskian terms the older child is able to

operate well within the toddler's ZPD.

Mixed-age play, for the older peer, also provides a context for the development

of pro-social behaviours including helping, teaching, sharing, explaining, and the

exchange of information (Katz et al., 1990). Each of these activities is thought to make

a valuable contribution to the development of more complex life skills (Katz et al.,

1990).

It was also noted in Chapter T w o (2.3) that Lougee et al. (1977) reported that in

mixed-age groups each partner was able to successfully incorporate both social and

behavioural adjustments to accommodate the other partner's activity.

7.21.6.2 Benefits for the Older Peers' Language and Communication Skill

Development

The preschool period coincides with some important developments in language

and the acquisition of communication skills. The task of sharing play materials and

playing with younger peers provides a context for the older children to practise the

semantic and pragmatic components of language. First, in semantic development,

preschoolers are likely to be acquiring words that specify relations between people, for

example using you and I. Secondly, they learn the words that specify relations between

objects and define spatial opposites - for example, tall and short, high and low (Shaffer,

1989). The preschool period is one in which children also discover the rules of

pragmatics of using language appropriately and effectively in communication with

others (Shaffer, 1989). Mixed-age play provides a context for the practice of these

newly acquired language skills.

Preschoolers are also becoming more skilled in communicating as they are able

first to recognise that their audience m a y require them to adjust their language and

secondly, to formulate the required modification to suit or fit that audience (Shatz,

1983; Warren & McCloskey, 1993). The 'reformulating of messages' is another

258


communication skill acquired and refined in the preschool period (Evans, 1985). Again

it is likely that this skill would be practised in the dyad situation.

Finally the content of preschoolers 'talk' while, as described by Lindfors (1980,

p. 151) as being "anchored in the situation of the moment" is also such that they can

speak about past experiences or future plans and events.

In summary there are a number of benefits for older children engaged in dyad

play with younger peers including some that enhance both socio-emotional and

language and communication skills development. Although there were no apparent

language or behavioural concessions made by the older peers in consideration of the

younger members of the dyad, an investigation into further effects of mixed-age play for

older peers would be informative.

7.22 CONCLUSIONS

This chapter contained a detailed discussion of the results - both those in

accordance and not in accordance with the predictions made by this researcher. Some

results were not consistent with this writer's expectations and explanations for these

have been offered. As the theories of Piaget and Vygotsky are central to this research,

results were explored mainly in the context of these two theoretical perspectives.

Relevant literature (reviewed in Chapter T w o ) was used to explain or extend

understandings arising from the results. Additional references were included where

appropriate.

The significant results shown in the current study whereby toddler participants

exhibited more frequent events of symbolic play and higher-level symbolic play in

Condition Three (dyad play) indicate that influences other than developmental

maturation are instrumental in the development of toddlers' symbolic play.

The results of the current enquiry, supplemented by the descriptive data, indicate

that mixed-age play, particularly dyad play (where all but four toddler's were paired

with a familiar, 'self-chosen' older peer), is a beneficial context for the younger player's

development of symbolic-play skills. Adoption and execution of this strategy - dyad

play incorporating a degree of structure and contrivance - requires the thoughtful

consideration of early-childhood professionals mindful of all aspects of very young

children's social and cognitive development.

259


A number of implications for programming in L D C C C s in ways that can support

the younger children's development of symbolic play were indicated in this chapter

together with some suggestions for future research.

260

Chapter 8: Conclusion

CHAPTER 8: CONCLUSION 8.1 OVERVIEW OF STUDY AND MAJOR FINDINGS

A n investigation into some of the influences on toddlers' emerging symbolic

play was the focus of this study. The project was undertaken in eight inner-city

L D C C C s in Sydney, Australia where the 24 participants (aged approximately from 18

to 30 months) had the opportunity to engage in free play with their older, preschool-

aged peers in either mixed-age play or dyad play. The mixed-age free-play session was

part of the regular L D C C C program and the current writer had no influence on that

condition.

The dyad-play condition was, however, structured in the following ways: the

three sets of play materials were supplied by the current researcher; and the toddler

participant's play partner was an older preschool-aged peer who was familiar and 'self-

chosen' by the toddler (that is, identified as a regular companion to the toddler in the

group mixed-age free-play session). Dyad play was conducted within, but a little aside

of, the children's main activity area.

The prediction was that toddler participants in these two conditions (mixed-age

play and dyad play) would exhibit symbolic play more frequently and at higher levels

than when engaged in play with their same-age peers.

The results supported this expectation. In dyad play (Condition Three) toddler

participants engaged in more frequent events of symbolic play and higher levels

symbolic play (according to Westby's [1991] Scale for Assessing Children's Pretend

Play) than in either of the other two conditions (same-age play or mixed-age play). The

dyad condition was also found to be an explanatory influence in the likelihood (or odds

ratio) for the youngest participants to exhibit more complex play (play that was 'greater

than that expected for their age') in two dimensions - 'thematic content' and 'self-other

relationships'. Finally dyad play (Condition Three) was found to be a predicting

variable for the likelihood for participants to exhibit complex symbolic play in both

'thematic content' (Dimension T w o ) and 'self-other relationships' (Dimension Four).

(Additionally first-born participants showed both more events of symbolic

activity and higher levels of symbolic play in mixed-age play - Condition Two.)

261


8.2 OTHER INFLUENCES

Additional explanatory influences found to be statistically significant in

toddlers' enhanced levels of symbolic play were the attendance of four or more days at

L D C C and the unsatisfactory provision of play materials. It is thought the older peers'

communicative behaviours m ay have served to 'tutor' or 'scaffold' the younger

children's play and activity. Additionally, the literature reviewed in Chapter T w o (2.4)

indicates that observational learning has a role in children's acquisition of symbolic-

play behaviour - the toddler participants in this study were observed to engage imitative

behaviour. Mixed-age play sessions in the L D C C C context offer opportunities for

younger children to observe others engaged in higher level or more complex symbolic

play.

8.3 THEORETICAL CONCLUSIONS

The results of this study confirmed that very young children generally follow a

predictable pathway as they develop the skills of symbolic play - an aspect of cognitive

development closely aligned with both social and language development. The four

dimensions of symbolic play as identified by Westby (1991) - 'decontextualisation',

'thematic content', 'organisation of themes' and 'self-other relationships' - are each

closely linked with areas of cognitive development. Whilst some parallels are both

expected and evident in the developmental pathway there exists too the opportunity for

each of these dimensions to be expressed and practised either with or ahead of the

others. In conclusion, very young children's symbolic play adheres closely to the

pathway defined by Westby (1991), whose Scale for Assessing Children's Pretend Play

appears to follow the Piagetian expectations; but opportunities for variability are also

recognised.

T w o premises central to Vygotskian theory employed in this enquiry are the

'zone of proximal development' and the notion of 'guided participation'. "Assisted

performance" (Bodrova & Leong, 1996, p. 35), a 'more-skilled other' or 'tuition' are

terms, indicating the involvement of another in a supportive role in a learning situation.

Bodrova and Leong describe 'assisted performance' as such that it "reveals the

behaviours that are on the verge of emerging" (1996, p.38). In the current enquiry,

when the child engaged in pretend-play activity alone, the emergent symbolic-play

skills m a y have remained unused or undiscovered.

262


8.4 FINAL CONCLUSIONS

As stated earlier, results showed that the toddler participants in this current study

benefited from their engagement, in both mixed-age play and dyad play, with their

preschool-aged peers. The dyad condition though, appears to be the most powerful

mixed-age context for the reasons described below.

First, the presence of an older peer engaged in play with a younger child

comprises a social context in which a number of elements or 'mechanisms' supportive

of the younger child's learning and skill acquisition are likely to occur. The

'mechanisms' engaged in by the older peer - 'the more skilled or accomplished other' -

include modelling and demonstrations of play behaviours. The older peers used

communicative behaviours that often served to 'tutor' or 'scaffold' the younger

children's play and activity. The social context presented in dyad play provided

opportunities for the Vygotskian notion of the 'zone of proximal development' to be

established and Rogoff s concept of "guided participation" (1990, p. 8) to occur.

In addition the dyad condition is likely to enhance interactional processes

supporting the development of: "joint involvement episodes" (Schaffer, 1996, p. 252)

and the 'joint action' and 'joint reference' postulated by Bruner (1975) as essential to

social-skill acquisition. Perhaps most importantly experience in dyadic situations

(viewed by Bronfenbrenner as a "basic building block of the microsystem", 1979, p. 56)

furthers the coordination of each partner's attention to constitute "joint attention"

(Tomasello, 1995, p. 107). Each of these 'mechanisms' or components of the dyad

social context, either singly or in conjunction with one another, have the capacity to

further the development of younger children's symbolic play.

The efficacy of the dyad condition was increased through the provision of

appropriate materials together with space and time for the engagement between the

older peer and the younger participant to occur and, in turn, their engagement with the

play materials.

Lastly, as mentioned in Chapter One (1.9), symbolic play is a precursor for

thematic social play - play which supports children's development in a range of areas

including perspective taking and impulse control (Wyver & Spence, 1995). Howes and

Matheson (1992, p. 969) found that toddlers who spent a larger proportion of time in

cooperative social pretend (symbolic) play showed earlier emergence of complex social

pretend (symbolic) play and, as preschoolers, spent more time engaged in such play.

263


The current author suggests that outcomes in keeping with those cited above are clearly

desirable and the dyad-play context, as implemented in the present study, is one method

that promises to contribute to these outcomes.

264

Bibliography

BIBLIOGRAPHY

265

Bibliography

BIBLIOGRAPHY This bibliography contains items specifically cited in the thesis and items to

which general reference was made.

Abramovitch, R., Corter, C, & Lando, B. (1979). Sibling interaction in the home. Child Development, 50, 997-1003.

Abramovitch, R., Corter, C , & Pepler, D. J. (1980). Observations of mixed-sex sibling dyads. Child Development, 51, 1268-1271.

Abramovitch, R., Corter, C , Pepler, D. J., & Stanhope, L. (1986). Sibling and peer interactions: A final follow-up and comparison. Child Development, 57, 217-229.

ABS. (1999). Childcare. Cat. No. 4402.0. Australian Bureau of Statistics: Canberra. ABS. (1999). Children, Australia: A social report. Cat. No. 4119.0. Australian Bureau

of Statistics: Canberra. ABS. (1998). Australia Now. Australian Bureau of Statistics: Canberra. Almy, M . (1984). A child's right to play. Young Children, 39 (4), 80. Andersson, B. (1989). Effects of public day-care: A longitudinal study. Child

Development, 60, 857-866. Aries, P. (1962). Centuries of childhood. NewYork: Knopf. Arnett, J. (1989). Caregivers in day-care centers: Does training matter? Journal of

Applied Developmental Psychology, 10, 541-552. Aureli, T., & Colecchia, N. (1996). Day care experience and free play behaviour in

preschool children. Journal of Applied Developmental Psychology, 17, 1-17. Aureli, T., & Procacci, M . A. (1992). Day-care experience and children's social

development. Early Child Development and Care, 83, 45-54. Austin, G. R. (1976). Early childhood education: An international perspective.

N e w York: Academic Press. Bailey, D. B., Burchinal, M . R., & McWilliam, R. A. (1993). Age of peers and early

childhood development. Child Development, 64, 848-862. Bailey, D. B., McWilliam, R. A., Ware, W . B., & Burchinal, M . R. (1993). Social

interactions of toddlers and preschoolers in same-age and mixed-age play groups. Journal of Applied Developmental Psychology, 14, 261-276.

Bandura, A. (1971). Analysis of modeling processes. In A. Bandura (Ed.), Psychological modeling: Conflicting theories. N e w York: Lieber-Atherton.

Bandura, A. (1977). Analysis of modeling processes. Social learning theory. Englewood Cliffs, NJ: Prentice-Hall.

Bandura, A. (1986 j. Social foundations of thought and actions: A social cognitive Theory. Englewood Cliffs, NJ: Prentice-Hall.

Bandura, A. (1989). Social cognitive theory. In Annals of Child Development, (Vol, 6) (1-60). Greenwich, Conn: JAI Press Inc.

Beizer, L., & Howes, C. (1992). Mothers and toddlers: Partners in early symbolic play: Illustrative study # 1. In C. Howes, O. Unger, & C. C. Matheson (Eds.), The collaborative construction of pretend: Social pretend play functions. (25-43).

N e w York: State University of N e w York Press. Belsky, J. (1988). The "effects" of infant day care reconsidered. Early Childhood

Research Quarterly, 3, 235-272. Belsky, J., & Steinberg, L. D. (1978). The effects of day care: A critical review. Child

Development, 49, 929-949.

266

Bibliography

Belsky, J., Garduque, L., & Hrncir, E. (1984). Assessing performance, competence and executive capacity in infant play: Relations to home environment and security of attachment. Developmental Psychology, 20 (3), 406-417.

Berk, L. E. (1994). Vygotsky's theory: The importance of make-believe play. Young Children, 50 (1), 291-296.

Berk, L. (1997). Child development. (4th ed.). M A : Allyn & Bacon. Biben, M., & Suomi, S. J. (1993). Lessons from primate play. In K. Macdonald (Ed.),

Parent-child play: descriptions and implications. (185-196). Albany, N e w York: University of N e w York Press.

Bjorklund, D. F. (1997). The role of immaturity in development. Psychological Bulletin, (122), 2, 153-169.

Blasco, P. M., Bailey, D. B. Junior., & Burchinal, M. A. (1993). Dimensions of mastery in same-age and mixed-age integrated classrooms. Early Childhood Research Quarterly, 8, 193-206.

Bloch, M . N. (1989). Young boys' and girls' play at home and in the community: A cultural-ecological framework. In M . Bloch, & A.D. Pellegrini (Eds.), The ecological contexts of children's play. (120-154). Norwood, NJ: Ablex.

Bodrova, E., & Leong, D. J. (1996). Tools of the mind: The Vygotskian approach to early childhood education. Englewood Cliffs, NJ: Prentice Hall.

Bornstein, M . H , Haynes, O. M., O'Reilly, A. W., & Painter, K. M. (1996). Solitary and collaborative pretense play in early childhood: Sources of individual variation in the development of representational competence. Child Development, 67, 2910-2929.

Bornstein, M . H., Haynes, O. M., Legler, J. M., O'Reilly, A. W., & Painter, K. (1997). Symbolic play in childhood: Interpersonal and environmental context and stability. Infant Behaviour & Development, 20 (2), 197-207.

Borstelmann, L. J. (1983). Children before psychology: Ideas about children from an antiquity to the late 1800s. In P. H. Mussen (Ed.), Handbook of child psychology (Vol, 1). (1-40). N e w York: Wiley.

Brody, G. H., & Stoneman, Z. (1981). Selective imitation of same-age, older and younger peer models. Child Development, 52, 717-720.

Bronfenbrenner, U. (1979). The ecology of human development. Cambridge, M A : Harvard University Press.

Bronfenbrenner, U. (1993). A new head start for Head Start. The 2nd National Head Start Research Conference. Washington, D C , November 4-7, 1993. (Condensed Version)

Brownwell, C. A. (1990). Peer social skills in toddlers: Competencies and constraints illustrated by same-age and mixed-age interaction. Child Development, 61, 838-

848. Bruner, J. (1972). The nature and uses of immaturity. American Psychologist, 27, 687-

708. Bruner, J. (1973). Organisation of early skilled action. Child Development, 44, 1-11. Bruner, J. (1974). Child's play. New Scientist, 62, 126-128. Bruner, J. (1975). The ontogenesis of speech acts. Journal of Child Language, 2,1-19. Bruner, J. (1977). Introduction. In B. Tizzard, & D. Harvey (Eds.), The biology of play.

(v-vi). London, UK: Spastics International Medical Publications. Bruner, J. (1983). Play, thought and language. Peabody Journal of Education, 60, 60-

68.

267

Bibliography

Bruner, J. (1984). Vygotsky's zone of proximal development: The hidden agenda. In B. Rogoff, & J. V. Wertsch (Eds.), Children's learning in the 'zone of proximal development.' (93-97). New Directions for Child Development, no. 23, San Francisco: Jossey-Bass.

Carpenter, M., Tomasello, M., & Savage-Rumbaugh, S. (1995). Joint attention and imitative learning in children, chimpanzees, and enculturated chimpanzees. Social Development, 4 (3), 217-237.

Case, R. (1985). Intellectual development: A systematic reinterpretation. N e w York: Academic Press.

Case, R. (1991). The mind's staircase. Hillsdale, NJ: Lawrence Erlbaum. Case, R., & Okamoto, Y. (1996). The role of central conceptual structures in the

development of children's thought. Monographs of the Society for Research in Child Development 61, (1-2, Serial N o 246).

Chaille, C , & Silvern, S. B. (1996). Understanding through play. Childhood Education, 72 (5), 274-277.

Christie, J. R, & Wardle, F. (1992). H o w much time is needed for play? Young Children, 47 (3), 28-32.

Cicirelli, V. G. (1976). Siblings teach siblings. In V. L. Allen (Ed.), Children as teachers: Theory and research on tutoring. (99-111). UK: Academic "Press.

Cohen, L., & Manion, L. (1980). Research methods in education. (2nd ed.). London, Great Britain: Croom Helm.

Connell, W . F. (1980). A history of education in the twentieth century world. Canberra, Australia: Curriculum Development Centre.

Dale, N. (1989). Pretend play with mothers and siblings: Relations between early performance and partners. Journal of Child Psychology & Psychiatry, 30 (5),

751-759. Damast, A. M., Tamis-LeMonda, C. S., & Bornstein, M . H. (1996). Mother-child play:

Sequential interactions and the relation between maternal beliefs and behaviours. Child Development, 67, YI52-1166.

Dansky, J. L., & Silverman, I. W . (1973). Effects of play on associative fluency in preschool children. Developmental Psychology, 9, 38-43.

Dasen, P. R., & de Ribaupierre, A. (1987). In A. Demetriou (Ed.), The neo-Piagetian theories of cognitive development: Toward an integration. (287-326).

Amsterdam, NL: Elsevier Science. de Lacey, P. R., & Cocking, R. R. (1974). A theoretical framework for the role of

guided play in early learning. The Annual Conference on Cognitive Development. Canberra, A C T , February, 1974.

de Lacey, P., Walker, S., & Nurcombe, B. (1999). (Eds.), Children of the dispossessed: Far-west preschoolers 30 years on. (2nd ed.). Stamford, CT: Ablex.

Demetriou, A. (1988). The neo-Piagetian theories of cognitive development: Toward an integration. Amsterdam, NL: Elsevier Science.

Dewey, J. (1916). Democracy and education: An introduction to the philosophy of education. N e w York: The Macmillan Company.

Diggle, P. J., Liang, K., & Zeger, S. L. (1994). Analysis of longitudinal data. Oxford:

Oxford University Press. Dockett, S. (1998). Constructing understandings through play in the early years.

International Journal of Early Years Education, 6 (1), 105-116. Donaldson, M . (1978). Children's minds. Glasgow, Great Britain: Fontana.

268

Bibliography

Dunham, P. J., & Moore, C. (1995). Current themes in research on joint attention. In C. Moore, & P. J. Dunham (Eds.), Joint attention: Its origins and role in development. (15-28). Hillsdale, NJ: Lawrence Erlbaum.

Dunn J., & Wooding, C. (1977). Play in the home and its implications for learning. In B. Tizard, & D. Harvey (Eds.), Biology of play. (45-58). London: SIMP.

Dunn, J., & Dale, N. (1984). "I a Daddy": 2-year-old's collaboration in joint pretend with sibling and with mother. In I. Bretherton (Ed.), Symbolic play: The development of social understanding. (249-270). N e w York: Academic Press.

Dunn, J., & Plomin, R. (1990). Separate lives: Why siblings are so different. New York: Harper Collins .

Dunn, L. (1993). Ratio and group size in day care programs. Child and Youth Care Forum, 22 (3), 193-226.

Dunn, L. M., & Dunn, L. M . (1981). Peabody picture vocabulary test revised. Circle Pines, M N : American Guidance Service.

Dunn, L., & Herwig, J. E. (1992). Play behaviours and convergent and divergent thinking skills of young children attending full-day preschool. Child Study Journal, 22, 23-35.

Dunn, L., Kontos, S., & Potter, L. (1996). Mixed-age interactions in family child care. Early Education and Development, 7 (4), 349-366.

Duveen, G., & Lloyd, B. (1988). Gender as an influence in the development of scripted pretend play. British Journal of Developmental Psychology, 6, 89-95.

Ebbeck, M . (1996).Children constructing their own knowledge. International Journal of Early Years Education, 4, (2), 5-27.

Elkind, D. (1981b). The hurried child. Boston, M A : Addison-Wesley Publishing. Ellis, S., Rogoff, B., & Cromer, C C . (1981). Age segregation in children's social

interactions. Developmental Psychology, 17, 399-407. Enz, B., & Christie, J. F. (1997). Teacher play interaction styles: Effects on play

behaviour and relationships with teacher training and experience. International Journal of Early Childhood Education, 2, 55-75.

Erikson, E. H. (1963). Childhood and society. (2nd ed.). N e w York: Norton & Co. Erikson, E. H. (1976). Play and actuality. In J. S. Bruner, A. Jolly, & K, Sylva (Eds.),

Play: Its role in development and evolution. (688-703). Aylesbury, GB: Hazell, Watson & Viney.

Evans, M.A. (1985). Self-initiated speech repairs: A reflection of communicative monitoring in young children. Developmental Psychology, 21, 365-371.

Falbo, T., & Polit, D. F. (1986). Quantitative review of the only child literature: Research evidence and theory development. Psychological Bulletin, 100 (2), 176-189.

Farver, J. M . (1992). Communicating shared meaning in social pretend play. Early Childhood Research Quarterly, 7, 501-516.

Farver, J. M., & Howes, C. (1993). Cultural differences in American and Mexican mother-child pretend play. Merrill-Palmer Quarterly, 3 (3), 345-358.

Farver, J. M., & Wimbarti, S. (1995). Paternal participation in toddlers' pretend play. Social Development, 4 (I), 17-31.

Fein, G. G. (1975). A transformational analysis of pretending. Developmental Psychology, 11 (3), 291-296.

Fein, G. G., & Apfel, N. (1979). Some preliminary observations on knowing and pretending. In N. R. Smith, & M . B. Franklin (Eds.), Symbolic functioning in childhood. (87-101). Hillsdale, NJ: Lawrence Erlbaum.

269

Bibliography

Fein G. G., & Fryer, M . G. (1995). Maternal contributions to early symbolic play competence. Developmental Review, 15, 367-381.

Feitelson, D., & Ross, G. S. (1973). The neglected factor - play. Human Development, 16, 202-223.

Fenson, L. (1984). Developmental trends for action and speech in pretend play. In I. Bretherton (Ed.), Symbolic play: The development of social understanding. (249-210). N e w York: Academic Press.

Fenson, L., & Ramsay, D. S. (1980). Decentration and integration of the child's play in the second year. Child Development, 51, 171-178.

Fiese, B. H. (1990). Playful relationships: A contextual analysis of mother-toddler interaction and symbolic play. Child Development, 61, 1648-1656.

Field, T. M . (1994). In H. Goelman, & E . V. Jacobs (Eds.), Children's play in childcare settings. (69-84). N e w York: State of N e w York Press.

Fisher, E. P. (1992). The impact of play on development: A meta-analysis. Play and Culture, 5, 159-181.

Fischer, K. W., & Farrar, M . J. (1987). Generalisations about generalisation: H o w a theory of skill development explains both generality and specificity. In A. Demetriou (Ed.), The neo-Piagetian theories of cognitive development: Toward an integration. (137-171). Amsterdam, NL: Elsevier Science.

Flavell, J. H., Miller, P. H., & Miller, S. A. (1993). Cognitive development. (3rd ed.). Englewood Cliffs, NJ: Prentice Hall.

Fossey, D. (1979). Development of the mountain gorilla (gorilla gorilla geringei): The first thirty-six months. In D. A. Hamberg, & E. R. McCoun (Eds.), The great apes. (139-186). Menlo Park, CA: Benjamin & Cummings.

Francis, P. L., & Self, P. A. (1982). Imitative responses of young children in day-care and home settings: The importance of the child-to-caregiver m.'io.Child Study Journal, 12 (2), 119-126.

French, D. C , Wass, G. A., Stright, A. L., & Baker, J. A. (1986). Leadership asymmetries in mixed-aged children's groups. Child Development, 57, 1277-

1283. Garvey, C. (1977). Play. Cambridge, M A : Harvard University Press. Garvey, C. (1990). Play. London: Fontana/Open. Garvey, C , & Kramer, T. L. (1989). The language of social pretend play.

Developmental Review, 9, 364-382. Giffen, H. (1984). The coordination of meaning in the creation of a shared make-believe

reality. In I. Bretherton (Ed.), Symbolic play: The development of social understanding. (73-101). N e w York: Academic Press.

Gleason, J. B. (1997). The development of language. (4th ed.). M A : Allyn & Bacon. Goffman, E. (1974). Frame analysis: An essay on the organisation of experiences.

Cambridge, M A : Harvard University Press. Goldman, J. A. (1981). Social participation of preschool children in same-versus mixed-

age groups. Child Development, 52, 644-650. Gowen, J. W . (1995). The early development of symbolic play. Young Children, 50 (3),

75-83. Haight, W., & Miller, P. J. (1992). The development of everyday pretend play:A

longitudinal study of mothers' participation. Merrill-Palmer Quarterly, 38, (3),

331-349. Haight, W . H., Masiello, T., Dickson, K. L., Huckeby, E., & Black, J. E. (1994). The

everyday contexts and social functions of spontaneous mother-child pretend

play in the home. Merrill-Palmer Quarterly, 40 (4), 509-522.

270

Bibliography

Haight, W . L., Parke, R. D., & Black, J. E. (1997). Mothers' and fathers' beliefs about and spontaneous participation in their toddlers' pretend play. Merrill-Palmer Quarterly, 43 (2), 271-290.

Halford, G. S. (1993). Children's understanding: the development of mental models. Hillsdale, NJ: Lawrence Erlbaum.

Hanna, E., & Meltzoff, A. N. (1993). Peer imitation by toddlers in laboratory, home and day-care contexts: Implications for social learning and memory. Developmental Psychology, 29 (4), 701-710.

Harms, T., Cryer, D., & Clifford, R. M . (1990). The infant-toddler environment rating scale. N e w York: Teachers College Press.

Hartup, W . W . (1976). Cross-age versus same-age peer interaction: Ethological and cross-cultural perspectives. In V. L. Allen (Ed.), Children as teachers: Theory and research on tutoring. (41-55). N e w York: Academic Press.

Hartup, W . W . (1983). Peer relations. In P. H. Mussen (Ed.), Handbook of child psychology>:Vol. 4. Socialisation, personality and social development. (4th ed.). (693-774). N e w York: Wiley.

Holloway, S. D., & Reichhart-Erickson, M . (1988). The relationship of day care quality to children's free-play behaviour and social problem-solving skills. Early Childhood Research Quarterly, 3, 39-53.

Holloway, S. D., & Reichhart-Erickson, M . (1989). Child-care quality, family structure and maternal expectations: Relationship to preschool children's peer relations. Journal of Applied Developmental Psychology, 10, 281-298.

Honig, A. S., & Hirallal, A. (1998). Which counts more for excellence in childcare staff years in service, education level or E C E coursework? Early Child Development and Care, 145, 31-46.

Howe, N., Moller, L., & Chambers, B. (1994). Dramatic play in day care: What happens when doctors, cooks, bakers, pirates and pharmacists invade the classroom? In H. Goelman, & E. V. Jacobs (Eds.), Children's play in child care settings. (102-118). N e w York: State of N e w York Press.

Howell, D. C. (1997). Statistical methods for psychology. (4th ed.). Boston, M A S S : Duxbury Press.

Howes, C. (1998). Child care caregiver sensitivity and attachment. Social Development, 7(l),26-36

Howes, C. (1980). Peer play scale as an index of complexity of peer interaction. Developmental Psychology, 16, 371-372.

Howes, C. (1985) Sharing fantasy: Social pretend play in toddlers. Child Development, 56, 1253-1258.

Howes, C. (1988). Relations between early child care and schooling. Developmental Psychology, 24(1), 53-57.

Howes, C. (1990). Can the age of entry into child care and the quality of child care predict adjustment in kindergarten? Developmental Psychology, 26 (2), 292-303.

Howes, C. (1991). A comparison of preschool behaviours with peers when children enrol in child-care as infants or older children. Journal of Reproductive & Infant Psychology, 9, 105-115.

Howes, C. (1997). Children's experiences in center-based child care as a function of teacher background and adultxhild ratio. Merrill-Palmer Quarterly, 43 (3), 404-

25. Howes, C , & Farver, J. (1987). Social pretend play in 2-year-olds: Effects of age of

partner. Early Childhood Research Quarterly, 2, 305-314.

271

Bibliography

Howes, C , Hamilton, C. E., & Matheson, C. C. (1994). Differential associations with aspects of the teacher-child relationship. Child Development, 65, 253-263.

Howes, C , & Hamilton, C. E. (1992). Children's relationships with caregivers: Mothers and child care teachers. Child Development, 63, 859-866.

Howes, C , & Matheson, C. C. (1992). Sequences in the development of competent play with peers: Social and pretend play. Developmental Psychology, 28 (5), 961-74.

Howes, C , & Hamilton, C. E. (1992a). Children's relationships with caregivers: Mothers and child care teachers. Child Development, 63, 859-866.

Howes, C , & Hamilton, C. E. (1992b). Teachers: Stability and concordance with parental attachments. Child Development, 63, 867-878.

Howes, C , Hamilton, C. E., & Matheson, C. C. (1994). Differential associations with aspects of the teacher-child relationship. Child Development, 65, 253-263.

Howes, C , Phillips, D. A., & Whitebook, M . (1992). Thresholds of quality: Implications for the social development of children in center-based child care. Child Development, 63, 449-460.

Howes, C , & Smith, E. W.> (1995). Relations among child care quality, teacher behaviour, children's play activities, emotional security and cognitive activity in child care. Early Childhood Research Quarterly, 10, 381-404.

Howes, C , & Stewart, P. (1987). Child's play with adults, toys and peers:An examination of family and child-care influences. Developmental Psychology, 23 (3), 423-430.

Howes, C , & Unger, O. (1989). Play with peers in child settings. In M. Bloch, & A.D. Pellegrini (Eds.), The ecological contexts of children's play. (104-117). Norwood, NJ: Ablex.

Howes, C , & Unger, O. (1992). Collaborative construction of social pretend play between toddler-age partners: Illustrative study #2. In C. Howes, O. Unger, & C. C. Matheson (Eds.), The collaborative construction of pretend. (45-54). New York: State University of N e w York Press.

Howes, C , Unger, O , & Seidner, L. Beizer. (1989). Social pretend play in toddlers: Parallels with social play and with pretend. Child Development, 60, 77-84.

Howes, C , & Unger, O. (1989). Play with peers in child settings. In M. Bloch, & A.D. Pellegrini (Eds.), The ecological contexts of children's play. (104-117). Norwood, NJ: Ablex.

Howes, C , Sakai, L., Shinn, M., Phillips, D., Galinsky, E., & Whitebook, M . (1995). Race, social class, and maternal working conditions as influences on children's development. Journal of Applied Psychology, 16, 107-124.

Hutt, C. (1971). Exploration and play in children. In R.E. Herron, & B. Sutton-Smith (Eds.), Child's Play. (231-251). N e w York: Wiley.

Johansson, B., & Sundell, K. (1989). Utvidgade syskongrupper - en atervandsgrand? Uppsala: Uppsala University, Department of Psychology. Cited in Sundell, K. (1994). Mixed-age groups in Swedish nursery and compulsory schools. School Effectiveness and School Improvement, 5, (4), 376-393.

Johnson, J. E. (1976). Relations of divergent thinking and intelligence test scores with social and nonsocial make-believe play of preschool children. Child

Development, 47, 1200-1203. Johnson, J. E. (1978). Mother-child interaction and imaginative behaviour of preschool

children. The Journal of Psychology, 100, 123-129. Johnson, J. E., & Ershler, J. (1982).Curricula effects of the play of preschoolers. In D. J.

Pepler, & K. H. Rubin (Eds.), The play of children: Current theory and

research. Basel: Karger.

272

Bibliography

Johnson, J. E., Ershler, J., & Lawton, J. T. (1982). Intellective correlates of

preschoolers' spontaneous play. The Journal of General Psychology, 106, 115-

Johnson, J. E. (1983). Context effects of preschool children's symbolic behaviour. The Journal of Genetic Psychology, 143, 259-268.

Johnson, J. E., & Yawkey, T. D. (1988). Play and integration. In T.D Yawkey, & J. E. Johnson (Eds.), Integrative processes and socialization: Early to middle Childhood. N e w Jersey: Lawrence Erlbaum.

Johnson, J. E. (1990). The role of play in cognitive development. In E. Klugman, & S. Smilansky (Eds.), Children's play and learning: Perspectives and policy implications. N e w York: Teachers College Press.

Kane, S. R., & Furth, H. G. (1993). Children constructing social reality: A frame analysis of social pretend play. Human Development, 36, 199-214.

Kaplan, G. M., & Conn, J. S. (1984). The effects of caregiver training on classroom setting and caregiver performance in eight community day care centres. Child Study Journal, 14 (2), 79-93.

Katz, L. G., Evangelou, D., & Hartman, J. A. (1990). The case for mixed-age grouping in early education. Washington, D.C.: N A E Y C .

Kerns, K. A., & Barth, J. M. (1995). Attachment and play: Convergence across components of parent-child relationships and their relations to peer competence. Journal of Social and Personal Relationships, 12 (2), 243-260.

Konner, M . (1975). Relations among infants and juveniles in comparative perspective. In M . Lewis, & I. Rosenblaum (Eds.), Friendship and peer relations. (99-129). N e w York: John Wiley & Sons.

Kontos, S., & Wilcox-Herzog, A. (1997). Teachers interactions with children: W h y are they so important? Young Children, 52 (2), 4-12.

Lamb, M . E. (1977b). Father-infant and mother-infant interaction in the first year of life. Child Development, 48, 167-181.

Lamb, M . E. (1978). Interactions between eighteen-month-olds and their preschool-aged siblings. Child Development, 49, 51-59.

Lamb, M . E., Sternberg, K. J., Knuth, N., Hwang, C. P., & Broberg, A. G. (1994). Peer play and nonparental care experiences. In H. Goelman, & E. V. Jacobs (Eds.), Children's play in child care settings. (37-52). N e w York: State of New York Press.

Lasater, C , & Johnson, J. E. (1994). Culture, play, and early childhood education. In L. Jaipaul, J. Roopnarine, J. E. Johnson, & F. H. Hooper (Eds.), Children's play in diverse cultures. (210-228). N e w York: State University of N e w York Press.

Lewis, M., Young, G., Brooks, J., & Michalson, L. (1975). The beginning of friendship. In M . Lewis, & I. Rosenblaum (Eds.), Friendship and peer relations. (27-66). N e w York: John Wiley & Sons.

Lewis, C , Freeman, N. H., Kyrriaakidou, C , Maarridaki-Kassotaki, K., & Berridge, D. M . (1996). Social influences on false belief access: Specific sibling influences or general apprenticeship? Child Development, 67, 22930-2947.

Lillard, A., & Curenton, S. (1999). D o young children understand what others feel, want and know? Young Children, 54 (5), 5257.

Lindfors, J. W . (1980). Children's language and learning. (2nd ed.). Englewood Cliffs

N.J.: Prentice-Hall. Lougee, M . D., Grueneich, R., & Hartup, W . W . (1977). Social interaction in same- and

mixed-age dyads of preschool children. Child Development, 48, 1353-1361.

273

Bibliography

Lyytinen, P. (1989). Effects of modelling on children's pretend play. Scandinavian Journal of Psychology, 30, 177-184.

Lyytinen, P. (1991). Peer interaction in children's dyad play. Early Child Development and Care, 71, 105-115.

Lyytinen, P. (1995). Cross-situational variation on children's pretend play. Early Child Development and Care, 105, 33-41.

Lyytinen, P., Poikkeus, A, -M., & Laakso, M . -L. (1997). Language and symbolic play in toddlers. International Journal of Behavioural Development, 21 (2), 289-302.

Mahoney, G., & Wheeden, C. A. (1999). The effect of teacher style on interactive engagement of preschool-aged children with special learning needs. Early Childhood Research Quarterly, 14 (1), 51-68.

Mandler, J. M., & McDonough, L. (1995). Long-term recall of event sequences in infancy. Journal of Experimental Child Psychology, 59, 457-474.

Marijnissen, R. H , & Seidel, M. (1984). Brueghel. N e w York: Harrison House. Matthew, W . S. (1977b). Modes of transformation in the initiation of fantasy play.

Journal of Developmental Psychology, 13, 212-216. McCartney, K., Scarr, S., Rocheleau, A., Phillips, D., Abbott-Shim, M., Eisenberg,

M., Keefe, N., Rosenthal, S., & Ruh, J. (1997). Teacher-child interaction and child-care auspices as predictors of social outcomes in infants, toddlers, and preschoolers. Merrill-Palmer Quarterly, 43 (3), 426-450.

McCune-Nicolich, L. M . (1977). Beyond sensorimotor intelligence: Assessment of symbolic maturity through analysis of pretend play. Merrill-Palmer Quarterly, 23, 89-99.

McCune-Nicolich, L. M . (1981). Toward symbolic functioning: Structure of early pretend games and potential parallels with language. Child Development, 52, 785-797.

McDonald, F. J. (1964). The influence of learning theories on education (1900-1950). In E. R. Hilgard (Ed.), Theories of learning and instruction: The sixty-third yearbook of the National Society for the Study of Education. (1-26). USA: The National Society for the Study of Education.

N C A C . (1993). Putting children first: Quality improvement and accreditation system handbook. Sydney, Australia: National Childcare Accreditation Council.

NcNeil, D. (1996). Epidemiological research methods. N Y : Wiley & Sons. Mead, G. H. (1932). Mind, self and society. Chicago, 111: University of Chicago Press. Meins, E., Fernyhough, C , Russell J., & Clark-Carter, D. (1998). Security of

attachment as a predictor of symbolic and mentalising abilities: A longitudinal

study. Social Development, 7(1), 1-24. Meltzoff, A. N. (1988). Nine-month-olds in immediate and deferred tests. Child

Development, 59, 217-225. Miller, G. A., Galnanter, E., & Pibram, K. H. (1960). Plans and the structure of

behaviour. N e w York: Holt. Moore, C , & Dunham, P. J. (1995). (Eds.), Joint attention: Its origins and role in

development. Hillsdale, NJ: Lawrence Erlbaum. Mounts, N. S., & Roopnarine, J. L. (1987). Social-cognitive play patterns in same-age

and mixed-age preschool classrooms. American Educational Research Journal,

24 (3), 463-476. NCOSS. (1997). Family friendly? You must be kidding. NCOSS child-care hotline.

Sydney, Australia: Council of Social Service N S W .

274

Bibliography

New, R. S. (1994). Child's play - una cosa naturale: A n Italian perspective. In J. L Roopnarine, J. E. Johnson, & F. H. Hooper (Eds.), Children's play in diverse cultures. (123-147). N e w York: State University of N e w York Press.

NICHD. Early Child Care Research Network. (1996). Characteristics of infant child care: Factors contributing to positive caregiving. Early Childhood Research Quarterly, 11, 269-306.

Nowak-Fabrykowski, K. (1994). Can symbolic play prepare children for their future? Early Child Development & Care, 102, 63-69.

O'Connell, B., & Bretherton, I. (1984). Toddler's play alone and with mother: The role of maternal guidance. In I. Bretherton (Ed.), Symbolic play: The development of social understanding. (337-368). N e w York: Academic Press.

Page, T. J., & Iwata, B. A. (1986). Interobserver agreement, history, theory and current methods. In A. Poling, & R. Fuqua (Eds.), Research methods in applied behaviour analysis: Issues and advance. (99-126). N e w York: Plenum.

Parten, M . B. (1932). Social participation among preschool children. Journal of Abnormal Psychology, 27, 243-269.

Pastor, D. L. (1981). The quality of mother-infant attachment and its relationship to toddlers' initial sociability with peers. Developmental Psychology, 17 (3), 326-335.

Peller, L. E. (1954). Libidinal phases, ego development, and play. Psychoanalytic Study of the Child, 9, 178-199.

Pepler, D. J., & Ross, H. S. (1981). The effects of play on convergent and divergent problem solving. Child Development, 52, 1202-1210.

Phillips, D., McCartney, K , & S. Scarr. (1987).Child-care quality and children's social development. Developmental Psychology, 23 (4), 537-543.

Piaget, J. (1954). The construction of reality in the child. N e w York: Basic Books. Piaget, J. (1959). The language and thought of the child. London: Routledge & Kegan

Paul. Piaget, J. (1962). Play, dreams and imitation in childhood. London: Routledge & Kegan

Paul. Plank, E. N. (1978). On development and education of young children: Selected papers

ofLilli E. Peller. N e w York: Philosophical Library. Plomin, R., & Daniels, D. (1987). W h y are children in the same family so different

from one another? Behavioural and Brain Sciences, 10, 1-60. Poling, A., Methot, L. L., & LeSage, M . G. (1995). Fundamentals of behaviour analytic

research. N e w York: Plenum Press. Rogoff, B. (1990). Apprenticeship in thinking: Cognitive development in social context.

N e w York: Oxford University Press. Rogoff, B., Mistry, J., Goncu, A., & Mosier, C. (1993). Guided participation in cultural

activity by toddlers and caregivers. Monograph of the Society for Research in Child Development, 58 (8, Serial No. 236).

Roopnarine, J. L., Almeduzzaman, M., Donnely, S., Gill, P., Mennis, A., Arky, L., Dingier, K , McLaughlin, M., & Talukder, E. (1992). Social-cognitive play behaviours and playmate preferences in same-age and mixed-age classrooms over a 6-month period. American Educational Research Journal, 29 (4), 757-

776. Rothstein-Fisch, C , & Howes, C. (1988). Toddler peer interaction in mixed-age groups.

Journal of Applied Developmental Psychology, 9, 211-218. Rubin, K. H. (1980). Fantasy play: Its role in the development of social skills and social

cognition. New Directions for Child Development, 9, 69-84.

275

Bibliography

Rubin, K. R., Maioni, T. L., & Hornung, M . (1976). Free play behaviours in middle-and lower-class preschoolers: Parten and Piaget revisited. Child Development 47t 414-419.

Rubin, K.H., Watson, K.S., & Jambor, T.W. (1978). Free-play behaviours in preschool and kindergarten children. Child Development, 49, 534-536

Rubin, K. H., Fein, G. G., & Vandenberg, B. (1983). Play. In P. H. Mussen (Ed.), Handbook of child psychology: Vol. 4. Socialisation, personality and social development. (4th ed). (693-774). N e w York: Wiley.

Rubenstein, J., & Howes, C. (1976). The effects of peers on toddler interactions with mother and toys. Child Development, 47, 597-605.

Rubenstein, J., & Howes, C. (1979). Caregiving and infant behaviour in day care and homes. Developmental Psychology, 15, 1-24.

Ruffman, T., Perner, J., Naito., Parkin, L., & Clements, W . A. (1998). Older (but not younger) siblings facilitate false belief understanding. Developmental Psychology, 3 (1), 161-174.

Rusher, A. S., Cross, D. R., & Ware, A. M. (1995). Infant-toddler play: Assessment of exploratory style and developmental level. Early Childhood Research Quarterly, 10, 297-315. '

Samuels, H. R. (1980). The effect of an older sibling on infant locomotor exploration of a new environment. Child Development, 51, 607-609.

Saracho, O. N. (1984). Construction and validation of the Play Rating Scale. Early Child Development and Care, 66, 45-61.

Scarr, S., Eisenberg, M., & Deater-Deckard, K. (1994). Measurement of quality in child care centres. Early Childhood Research Quarterly, 9, 131-151.

Schaffer, R. H. (1996). Joint involvement episodes as context for development. In H. Daniels (Ed.), An introduction to Vygotsky. (251-280). London: Routledge.

Schwarz, J. C , Strickland, R. G., & Krolick, G. (1974). Infant day care: Behavioural effects at preschool age. Developmental Psychology, 10 (4), 502-506.

Shaffer, D. R. (1989). Developmental psychology childhood and adolescence. (2° ed.). Pacific Grove, CA: Brooks/Cole.

Shatz, M., & Gelman, R. (1973). The development of communication skills: Modifications in the speech of young children as a function of listener. Monographs of the Society of Research in Child Development, 38 (Serial No. 152).

Shatz, M . (1983) Communication. In P. H. Mussen (Ed.). Handbook of child psychology: Vol. 3. Cognitive Development. N e w York: Wiley

Schiller, W . (2000). Arts, thinking, young children?: An overview. In W . Schiller, (Ed.), Thinking through the arts. Amsterdam, NL: Harwood Academic.

Schiller, W., & Veale, A. (1989). An integrated expressive arts program. Watson,

Australia: AECA. Siegel, S. (1956). Nonparametric statistics for the behavioural sciences. N e w York:

McGraw-Hill. Siegel, S., & Castellan, J. N. (1988). Nonparametric statistics for the behavioural

sciences. (2nd ed.). N e w York: McGraw-Hill. Siegler, R. S. (1991). Children's thinking. (2nd ed.). Upper Saddle River, NJ: Prentice-

Hall. Singer, J. (1994). Imaginative play and adaptive development. In J. H. Goldstein (Ed.),

Toys, play, child development. (6-26). Cambridge, England: Cambridge

University Press.

276

Bibliography

Slade, A. (1987a). A longitudinal study of maternal involvement and symbolic play during the toddler period. Child Development, 58, 367-375.

Slade, A. (1987b). Quality of attachment and early symbolic play. Developmental Psychology, 23 (1), 78-85.

Smilansky, S. (1968). The effects of sociodramatic play on disadvantaged preschool children. N e w York: Wiley.

Smilansky, S., & Shefataya, L. (1990). Facilitating play: A medium for promoting cognitive, socio-emotional and academic development in young children. Gaithersburg, M D : Psychosocial and Educational Publications.

Smith, P. K., & Vollstedt, R. (1985). On defining play: An empirical study of the relationship between play and various criteria. Child Development, 56, 1042-1050.

Spearritt, P. (1979). Child care and kindergartens in Australia, 1890-1975. In P. Langford, & P. Sebastian (Eds.), Early childhood education and care in Australia. (10-38). Melbourne: Harcourt Brace Jovanovich

Stambak, M., & Sinclair, H. (1993). Pretend play among 3-year-olds. (H. & M . Sinclair, Trans.) Hillsdale, NJ: Lawrence Erlbaum. (Original work published 1990.)

Stambak, M., Barriere, M., Bonica, L., Maisonnet, R., Musatti, T., Rayna, S., & Verba, M . (1983), Les bebes entre eux. Paris: Presses Universitaires de France. Cited in Stambak, M., & Sinclair, H. (1993). Pretend play among 3-year-olds. (H. & M. Sinclair, Trans.) Hillsdale, NJ: Lawrence Erlbaum. (Original work published 1990.)

State Government of N e w South Wales (1996). Centre based and mobile child care services regulations (no. 20). Under the Children's (Care and protection) Act, 1987. Gazette no. 125. Sydney, Australia: State Government of N S W .

Steele, C , & Hrncir, E. (1985). Adult suggestion: An enabler for pretend play in two-year-olds. In J. L. Frost, & S. Sunderlin (Eds.), When children play. Proceedings of the International Conference on Play and Play environments. M D : Association for Childhood Education International.

Sundell, K. (1994). Mixed-age groups in Swedish nursery and compulsory schools. School Effectiveness and School Improvement, 5, (4), 376-393.

Sylva, K., Roy, C , & Painter, M . (1980). Childwatching at playgroup and nursery school. U K : Grant Mclntyre.

Tabachnick, B. G., & Fidell, L. S. (Eds.), (1996). Using multivariate statistics. (3rded.).

N Y : Harper Collins. Tamis-LeMonda, C. S., Chen, L. A., & Bornstein, M . H. (1998). Mothers' knowledge

about children's play and language development: Short-term stability and interrelations. Developmental Psychology, 34 (1), 115-124.

Tegano, D. W., & Burdette, M . P. (1991). Length of activity periods and play behaviours of preschool children. Journal of Research in Childhood Education,

5, 93-98. Tegano, D. W., Lookabaugh, S., May, G. E., & Burdette, M . P. (1991). Constructive

play and problem solving: The role of structure and time in the classroom. Early Child Development and Care, 68, 27-35.

Teti, D. M., Bond, L. A., & Gibbs, E. D. (1988). Mothers, fathers and siblings: A comparison of play styles and their influence upon infant cognitive level. International Journal of Behavioural Development, 11 (4), 415-432.

Thomas, A., Chess, S., & Birch, H. G. (1968). The origins of personality. Scientific

American, 223 (2), 102-109.

277

Bibliography

Thut, I. N., & Adams, D. (1964). Educational patterns in contemporary societies. USA: McGraw-Hill.

Tomasello, M., Savage-Rumbaugh, S., & Kruger, A. C. (1993). Imitative learning of actions on objects by children, chimpanzees and enculturated chimpanzees. Child Development, 64, 1688-1705.

Tomasello, M . (1995). Joint attention as social cognition. In C. Moore, & P. J. Dunham (Eds.), Joint attention: Its origins and role in development.(103-129). Hillsdale, NJ: Lawrence Erlbaum.

Trawick-Smith, J. (1991). The significance of toddler pretend play in child care. Early Child Development and Care, 68, 11-18.

Tudge, J. R. H , & Winterhoff, P. (1993). Vygotsky, Piaget, and Bandura: Perspectives on the relation between the social world and cognitive development. Human Development, 36, 61-81.

Vandell, D. Lowe., & Powers, C. P. (1983). Day care quality and children's free play activities. American Journal of Orthopsychiatry, 53 (3), 493-500.

Vandenberg, B. (1981). Play: Dormant issues and new perspectives. Human Development, 24, 357-365.

Van Hoorn, J., Noorot, P., Scales, B., & Alward, K. (1993). Play at the center of the curriculum. N e w York: Macmillan.

Vygotsky, L. S. (1962). Thought and language. Cambridge, M A : MIT Press. Vygotsky, L. S. (1966). Play and its role in the mental development of the child. Soviet

Psychology, 12 (6), 62-76. Vygotsky, L. S. (1978). Mind in society: The development of the higher psychological

processes. Cambridge, M A : Harvard University Press. Vygotsky, L. S. (1987). The collected works ofL S. Vygotsky: Vol. 1. Problems of

general psychology. (R. W . Riever & A. S. Carton, Eds., N. Mimick, Trans.)

N Y : Plenum. Wangmann, J. (1995). Towards integration and quality assurance in children's

services. Melbourne: AIFS. Warren, A. R., & McCloskey, L. A. (1993). Pragmatics: Language in social contexts. In

J. Berko Gleason (Ed.), The development of language. (3rd ed.). New York:

Macmillan. Weininger, O. (1986). "What if" and "as if": Imagination and pretend play in early

childhood. International Journal of Early Childhood, 1, (22-28). Weisner, T.S. (1989). Comparing sibling relationships across cultures. In P. G. Zukow.

(Ed.), Sibling interaction across cultures: Theoretical and methodological

issues. (11-25). N e w York: Springer-Verlag. Werner, H., & Kaplan, B. (1963). Symbol formation. New York: Wiley. Wertsch, J. (1984). The zone of proximal development: Some conceptual issues. In B.

Rogoff, & J. V. Wertsch (Eds.), Children's learning in the 'zone of proximal development'. New Directions for Child Development, no. 23, San Francisco,

CA: Jossey Bass. Westby, C. E. (1980). Assessment of cognitive and language abilities through play.

Language, Speech and Hearing Services in Schools, 11, 154-168. Westby, C. (1991). A scale for assessing children's pretend play. In C. E. Shaefer, K,

Gitlin, & A. Sandgrund (Eds.), Play diagnosis and assessment. (131-161). New

York: John Wiley & Sons. Whiting, B. B., & Whiting, J. W . M. (1975). Children of six cultures. Cambridge, M A :

Harvard University Press.

278

Bibliography

Winnicott, D. W . (1965). The maturational processes and the facilitating environment. London: Hogarth Press & the Institute of Psycho Analysis.

Winnicott, D. W . (1971). Playing and reality. London: Tavistock. Wyver, S. R., & Spence, S. (1995). Cognitive and social play of Australian

preschoolers. Australian Journal of Early Childhood, 20 (2), 42-46. Wyver, S. R., & Spence, S. H. (1999). Play and divergent problem solving: Evidence

supporting a reciprocal relationship. Early Education & Development, 10 (4), 419-444.

Youngblade, L. M., & Dunn, J. (1995). Individual differences in young children's pretend play with mother and siblings: Links to relationships and understanding of other people's feelings and beliefs. Child Development, 66, 1472-1492.

279

Appendices

APPENDICES

280

Appendices

Appendix Ai

Excerpt from: Westby, Carol, E. (1991). A scale for .assessing children's pretend play. In C E . Schaefer, K, Gitlin, & A. Sandgrund. (Eds.), Play diagnosis and assessment. (131-161). N e w York: John Wiley & Sons.

Symbolic Level

Age in months

Bl

17-19

BI1

19-22

BIO

24

BIV

30

BV

36

BVI

36-42

Decontextualisation (What props are used in the play?)

Child exhibits internal mental representation Pretend, using lifelike props

As for level B I.

As for level B I

As for level B I

As for Level B I

Carries out pretend

activities with replica toys. Uses one object to represent another.

Thematic content (What schemas/scripts does the child represent?)

Familiar everyday activities, (e.g. eating, sleeping)

Activities of familiar others (e.g. cooking, reading, cleaning)

As for level B II

Represents personally experienced events that occur less frequently.

Compensatory play: Reenacts experienced events, but modifies original outcomes.

Observed events

(i.e. events in which child was not an active participant - e.g. police officer, schemas from T V shows)

Organisation of themes (How coherent and logical are the child's schemas and scripts?)

Short, isolated schemas (single

pretend actions)

Short, isolated schema comb inations (child combines two actions or toys in pretend; e.g. rocking doll and putting it to bed)

Elaborated single schemas (represents daily experiences with details, e.g. puts lids on pan, puts pan on stove)

Evolving episode sequences (e.g. child mixes cake, bakes it, serves it washes dishes)

As for Level B IV

As for Level BIV

Self-other relationships (What roles does the child take and give to toys and other people?)

Self as agent (autosymbolic or self-representational play; e.g. child pretends to go to sleep, to eat

Child acts on doll (doll is passive recipient of action) Child performs pretend actions on more than one object or person

As for level B II

Talks to doll

Transforms self into role.

Engages in associative play

Uses doll or

puppet as participant in play: Child talks

for doll. Reciprocal role taking, child assigns roles to other children.

281

Appendices

Appendix Aii

'Expected' Levels of Dimensions of Symbolic Play for Participants According to their

Age Group and the Westby (1991) Scale for Assessing Children's' Pretend Play

Dimensions of symbolic play

Age of participant in months

17-19 mths

20-22 mths

23-26 mths

27-31 mths

1. Decontextualisation

Range 1

1

1

1

1

2. Thematic content

Range 1-3

1

2

2

3

2. Organisation of themes Range 1-4

1

2

3

4

4. Self-other

relationships Range 1-3

1

2

2

3

282

Appendices

Appendix Bi Initial letter to centre Director

Institute of Early Childhood, Macquarie University, Ryde, N S W 2109 12th February 1997

Dear

Thank you for expressing interest in m y P h D research project. With reference to this project a copy of each of the following is attached for your perusal and information:

- parent information letter - parent consent form

- parent questionnaire (for parents of 'targeted' toddlers and dyad older peers only)

- staff member information letter - staff member consent form - staff member questionnaire/qualification scale

Towards the end of 1996 I successfully conducted a pilot study. Five centres were involved and it created the opportunity to fine-tune documentation systems as well as methodology. Just to alert you to what centre participation involves - I would like to be present for several (depending upon the number of participating toddlers) mixed-age outdoor free play sessions and several same-age outdoor free play sessions. The third condition is a structured dyad session where a participating toddler is paired with a 'chosen' older peer - I bring materials for this session. Nine, ten-minute segments of play (three in each condition) for each participating toddler are videotaped. N o special conditions or materials are needed and I would expect to be present for about an hour on perhaps six or seven occasions over approximately three weeks. M y attendance depends largely upon that of the 'targeted' children. During the pilot study I implemented a system for both distributing and collecting consent forms, which proved to be a welcome release for staff as it relieves them from handling and collecting consent forms. Perhaps w e could discuss this in more detail when w e meet. If your centre does decide to participate I would appreciate being able to visit, perhaps on one or two occasions - before any actual project work - so the children become used to the camera and me. The camera is small and about as unobtrusive as they come but I imagine, initially, both it and I may cause some interest! I realise that m y proposal will need to be approved by the centre's 'governing body' and to that end have enclosed information detailing the ethical aspects of the study. Ethics approval for this research project has already been granted by the University of Wollongong H u m a n Research Ethics Committee and K U Children's Services. Please, if you have any questions, contact m e on 9850 9832 any time. I'd also be delighted to speak with staff and/or parents re the project if or whenever you perceive

this as appropriate. Best regards,

Helen Kowalski

283

Appendices

Appendix Bii Letter to Long-Day Child-Care Centre Staff.

Study examining the development of pretend play in toddlers. May 1997 Information for centre staff. Helen Kowalski - P h D student, Graduate School of Education, University of Wollongong. Supervisors: Dr Philip de Lacey, Dr Grace Masselos, Dr Shirley Wyver

This project is being conducted as part of my PhD research study supervised by Dr Philip de Lacey and Dr Grace Masselos in the Graduate School of Education at the University of Wollongong and Dr Shirley Wyver from the Institute of Early Childhood at Macquarie University. M y area of study is the social pretend play of toddlers. As a member of staff you will have seen that simple pretend play is much in evidence at this age. This early pretend play is thought to be an important forerunner of the more complex types of play that develop later in childhood. A substantial amount of research supports the notion that children who are 'good players' experience enhanced cognitive, language and social skill development. The current project will examine the conditions under which toddlers are most likely to exhibit early pretend play. The procedure involves the researcher video-taping toddlers aged 18-30 months, with same-age peers and older peers aged 48-60 months engaged in free play, both indoors and outdoors, in their long-day child-care centre. Nine, ten-minute segments of participating toddlers' play will be video-taped over a period of approximately two weeks. Care will be taken that the observing and video-taping are conducted in a manner, which ensures minimal disruption to the children and environment. Children will not be withdrawn from their routine activities. The videos and field notes will subsequently be coded for pretend play content and analysed to identify what factors influenced the development of this play. It is proposed that factors, such as play materials and other children's suggestions will be seen to influence the development of early pretend play. A pilot study was successfully conducted from October to December last year. Staff members will be asked to complete a brief questionnaire but I would like to stress that there is no need for m e to obtain any personal details other than your first name and your qualifications. Information collected is completely confidential. N o individual names, videotapes or field notes will be seen by anyone other than the researcher, the supervisors and a research assistant required to establish inter-relater reliability. N o staff member will be asked to participate if permission is not obtained. If at any stage you wish to withdraw from the study you are free to do so. There will be no need for you to provide a reason and there will be no

penalty. The researcher will complete, in collaboration with the centre director, a brief environment rating scale - this will be held in confidence and safe storage by both the centre and the

researcher. If you feel you would be willing to participate in this research your co-operation would be both valued and appreciated. To enable participation please fill in the attached Consent Form and return the top section, in the stamped addressed envelope provided to Helen Kowalski — this portion will be held by the researcher. (A box marked 'consent form collection' has been provided next to the sign-in book at the centre). The other section should be retained by you for your records. (Consent forms and the video-tapes will be held in safe custody, by the researcher, for at least ten years). Your decision to participate or not participate in this study will not in any

way influence your position or status in the centre. At the conclusion of the project the results of the study together with a discussion of these will be distributed to parents and centre staff. If you require any further information regarding the project, please do not hesitate to call m e on 9850 9832 anytime. If you have any enquiries regarding the conduct of the research please contact the secretary of the University of

Wollongong H u m a n Research Ethics Committee on 042 214457.

Yours sincerely,

Helen Kowalski

284

Appendices

Appendix Biii Staff consent form to participate in study.

Staff Member Consent Form Please complete and return this portion, in the envelope provided.

I consent to participate in the study examining the influences upon the pretend play of toddlers.

Signed Date

If you change your mind about this study, you may withdraw at any time. You will not be required to give a reason and there will be no penalty.

One copy of the consent form will be kept by the researcher. Only individuals with direct involvement in this project will have access to this form. The other copy is to be retained by you.

The ethical aspects of this study have been approved by Wollongong University Ethics Review Committee. If you have any enquiries regarding the conduct of the research please contact the secretary of the University of Wollongong Human Research Ethics Committee on 042 214457.

Please cut here and retain this portion for your own records

Staff Member Consent Form

I consent to participate in the study examining the influences upon the pretend play of toddlers.

Signed Date


One copy of the consent form will be kept by the researcher. Only individuals with direct involvement in this project will have access to this form. The other copy is to be

retained by you.

The ethical aspects of this study have been approved by Wollongong University Ethics Review Committee. If you have any enquiries regarding the conduct of the research please contact the secretary of the University of Wollongong Human Research Ethics

Committee on 042 214457.

285

Appendices

Appendix Biv Letter of thanks to staff accompanying Staff Questionnaire

Institute of Early Childhood, Macquarie University, Ryde, N S W 2109 29th April 1997

Dear

Thank you for consenting to participate in my study - your decision was really appreciated. I would also like to thank you for being so supportive and enthusiastic during the current video-taping - you certainly make m e feel welcome and at ease in your centre. I have enclosed the staff questionnaire referred to in the information letter. This questionnaire is confidential and does not require any personal details. Your responses to the questions are an important component of the study. O n completion of the questionnaire please return it to m e in the enclosed envelope. (You'll find I've placed a collection box for these questionnaires in the staff room.) At the conclusion of the project the results of the study together with a discussion of these will be distributed to all staff. Please do not hesitate to call m e on 9850 9832 if you have any questions.

Thank you again for your participation. Yours faithfully,

Helen Kowalski

286

Appendices

Appendix Bv Staff Questionnaire

May 1997 CONFIDENTIAL

A study to identify influences on the development of toddlers' simple social pretend play.

Staff questionnaire Your participation in this research and your co-operation in completing this

questionnaire about children's development is both valued and appreciated. Please indicate your response in the spaces provided in the following manner:

strongly disagree disagree neither agree nor

1 2 3

1. Sound nutrition is essential for children's development.

2. It is essential for young children to experience free play.

3. Daily gross motor activity is not necessary for children's development and well-being.

disagree

1

1

1

2

2

2

agree

4

3

3

3

strongly agree

5

4 5

4 5

4 5

4. It is not essential to plan children's play experiences. 1 2 3 4 5_

5. Please describe what indicates to you that children are engaged in pretend play.

6. Please list what you consider to be 5 major influences on children's play.

7. Play positively influences children's cognitive development. 1 2 3 4 5_

8. All children benefit from taking part in mixed-age group free play sessions. 1 2 3 4 5_

9. Sleeping, not just resting or a quiet time is essential for children's well-being. 1 2 3 4 5

Thank you for taking the time to complete this questionnaire. At the conclusion of the project the results of the study together with a discussion of these will be distributed to centre staff.

If you require any further information regarding the project, please do not hesitate to call m e on 9850 9832 anytime. If you have any enquiries regarding the conduct of the research please contact the secretary of the University of Wollongong H u m a n Research Ethics Committee on 042 214457.

Yours sincerely,

Helen Kowalski Please turn over

287

Appendices

Appendix Bvi Staff Training and qualifications. C O N F I D E N T I A L Categories for specialised qualifications in Early-Childhood education. Please document highest qualification held.

Specialised qualifications in E C E 1. N o formal qualification

2. In-service workshops held at centre

3. Workshops in the community e.g. those held by local council

4. Workshops at professional meetings e.g. TEC Continuing Education course

5. High school courses e.g. year 10 or equivalent work experience placement

6. Private college courses e.g. nanny or child care course at, for example, The Australian Academy

7. T A P E Child Care Certificate Associate Diploma of Child Studies

8. Registered nurse

9. Mothercraft nurse

10. 3-year degree in Early-Childhood

Education

11.4-year degree in Early-Childhood

Education

12. Graduate diploma in Early-Childhood Education

13. Graduate degree in Early-Childhood

Education

14. Other

Response

Adapted from Howes, Carollee., & Smith, Ellen, W . (1995). Relations among child-care quality, teacher behaviour, children's play activities, emotional security and cognitive activity in child-care. Early Childhood Research, 10. 381-404

288

Appendices

Appendix Ci Parent information letter. Study examining the development of pretend plav in toddlers. February 1997 Parent information

Helen Kowalski - P h D student, Graduate School of Education, University of Wollongong. Supervisors: Dr Philip de Lacey, Dr Grace Masselos, Dr Shirley Wyver.

This study is being conducted as part of m y P h D research study supervised by Dr Philip de Lacey and Dr Grace Masselos in the Graduate School of Education at the University of Wollongong and Dr Shirley Wyver from the Institute of Early Childhood at Macquarie University. M y area of study is the social pretend play of toddlers. As parents you will have seen that simple pretend play is much in evidence at this age. This early pretend play is thought to be an important forerunner of the more complex types play that develop later in childhood. A substantial amount of research supports the notion that children who are 'good players' experience enhanced cognitive, language and social skill development. The current project will examine the conditions under which toddlers are most likely to exhibit early pretend play. The procedure involves the researcher video-taping toddlers aged 18-30 months engaged in free play with both same-age peers and older peers aged 48-60 months, both indoors and outdoors, in their long-day care centre. A brief task will be administered to the preschool-age participants, by the researcher, to gain an indication of receptive and expressive language skills. Nine, ten-minute segments, of each participating toddlers' play will be video-taped over a period of approximately two weeks. Attendance patterns and age will determine which toddlers participate. Care will be taken that the observing and video-taping are conducted is a manner, which ensures minimal disruption to the children and environment. Children will not be withdrawn from their routine activities. The video-tapes and field notes will subsequently be coded for pretend play content and analysed to identify what factors influence the development of the play. It is proposed that factors, such as play materials and other children's suggestions will be seen to influence the development of early pretend play. A pilot study was successfully conducted from October to December in 1996. Parents of participating toddlers and older peer partners will be asked to complete a brief questionnaire but I would like to stress that there is no need for m e to obtain any personal details regarding the child's family. All information collected is completely confidential. N o individual names, video-tapes or field notes of the children will be seen by anyone other than the researcher, the supervisors and a research assistant required to establish inter-relater reliability. Participants are selected according to age and attendance patterns. Generally as few as five or six children in each centre fit the participant profile. However, as they are observed in a mixed-age setting it is necessary to obtain parental consent for all children w h o attend the centre to take part in the project. Nevertheless, no child will be asked to participate if parental permission is not obtained. If at any stage you wish to withdraw your child from the study you are free to do so. There will be no need for you to provide a reason and there will be no penalty. If you feel you would like your child to participate in this research your co-operation would be both valued and appreciated. T o enable your child to participate please fill in the attached Consent Form and return the top section, in the envelope provided, to Child-care Centre - this portion will be held by the researcher. (A box marked 'consent form collection' has been provided next to the sign-in book in the centre). The lower section should be retained by you for your records. (Consent forms and video-tapes will be held by the researcher in safe storage for at least ten years). Your decision for your child to participate or not participate in this study will not in any way influence the services provided to your family by the long-day care centre. At the conclusion of the project the results of the study together with a discussion of these will be distributed to parents and staff members. If you require any further information regarding the project, please do not hesitate to call m e on 9850 9832 anytime. If you have any enquiries regarding the conduct of the research please contact the secretary of the University of Wollongong H u m a n Research Ethics Committee on 042 214457.

Yours sincerely,

Helen Kowalski

289

Appendices

Appendix Cii Parent consent form for child to participate in study.

Parent Consent Form

I give my consent for my child to participate in the study examining the social pretend play of toddlers.

Signed Date


One copy of the consent form will be kept by the researchers. Only individuals with direct involvement in this project will have access to this form. The other copy is to be retained by you.

The ethical aspects of this study have been approved by Wollongong University Ethics Review Committee. If you have any enquiries regarding the conduct of the research please contact the secretary of the University of Wollongong Human Research Ethics Committee on 042 214457.

Please cut here and keep one copy for your records

Parent Consent Form

I give my consent for my child to participate in the study examining the social pretend play of toddlers.

Signed Date


One copy of the consent form will be kept by the researchers. Only individuals with direct involvement in this project will have access to this form. The other copy is to be

retained by you.

The ethical aspects of this study have been approved by Wollongong University Ethics Review Committee. If you have any enquiries regarding the conduct of the research please contact the secretary of the University of Wollongong Human Research Ethics

Committee on 042 214457.

290

Appendices

Appendix Ciii

Letter of thanks to parents of participating toddlers

Institute of Early Childhood, Macquarie University, Ryde, N S W 2109 2nd June 1997

Dear Parent/s,

Thank you for consenting for your child, , to participate in my study. Your decision was really appreciated. The enclosed parent questionnaire, referred to in the information letter, is an important part of the project. It is confidential and does not require any personal details regarding your family. O n completion of the questionnaire please return it to m e in the enclosed stamped addressed envelope. At the conclusion of the project the results of the study together with a discussion of these will be distributed to parents. Please do not hesitate to call m e on 9850 9832 if you have any questions. Thank you again for your participation. Yours faithfully,

Helen Kowalski

291

Appendices

Appendix Civ: Parent Questionnaire j a n 1997

CONFIDENTIAL A study to identify some influences on the development of toddlers' simple social pretend play.

Parent questionnaire. Your participation in this research and your co-operation in completing this questionnaire is both valued and appreciated. Please write mUhe spaces indicated and return in the enclosed SAE.

1. What reasons led you to place

in long-day child care?

2. W h y did you choose long-day child-care centre rather than any other?

3. Please describe briefly what development you would expect to see (occur) in your child, in the next six month, in each of the following areas: - cognitive or intellectual - physical - emotional - social - other

4. What language is predominantly spoken at

home?

292

Appendices

5. Please list the birth dates and sex of all the children in your family (names are not required).

6. As part of the study I would like to know what sort of experiences children have during the weekends. Please choose one weekend leisure activity that your child often engages in and describe it briefly. This activity could be with parents, siblings, friends or others.

7. Age of parent/guardian. (1)

Relationship to child.

8. Age of parent/guardian. (2)

Relationship to child.

9. Level of education achieved by parent/guardian (1).

Present occupation.

10. Level of education achieved by parent/guardian (2).

Present occupation.

Under 21 years 21-30 years 31-40 years 41 years and over

Under 21 years 21-30 years

41 years and over

Completed Year 10 Completed Year 12 Completed a T A F E course Completed a University degree Other

Completed Year 10 Conmleted Year 12 Completed a T A F E course Completed a University degree Other

** I would be very interested in reading any comments you would like to add.

293

Appendices

Appendix D

Summary of PPVT-R (Form M ) Scores

Characteristics of Preschool-Aged Participants

Participant

Id

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

Sex

F

F

M

F

F

F

F

F

F

M

F

M

F

F

M

F

M

F

F

F

F

M

M

M

Age in months

45

47

47

48

48

48

49

50

51

51

52

52

53

54

54

55

55

56

56

56

59

61

62

63

PPVT-R

Raw score

50

53

51

56

34

56

47

64

50

68

59

58

85

66

56

54

68

55

54

54

32

73

77

41

PPVT-R

Scaled score

113

110

111

113

100

113

100

118

104

123

111

110

128

116

105

100

118

101

100

100

72

117

119

80

*18 and 19 represent the same child - indicated by two participants as preferred familiar

older peer. **21 and 24 represent older peers for whom English was a second language.

294

Appendices

Appendix E

Environment Rating Scale. CONFIDENTIAL Adapted from: Harms, T., Cryer, D., & Clifford, R. (1987). The infant and toddler environment rating scale.

N e w York: Teachers College Press.

Please circle the appropriate rating and write any additional comments in the space provided. A. Furnishings for children - Furnishings for routine care (feeding, sleeping and storage of child's possessions).

Inadequate Minimal Adequate Good Excellent ... 1 2 3 t

Insufficient number of pieces of furniture for feeding and sleeping routines (e.g. each child present does not have own cot or bed). Routine care furniture is generally in poor repair (e.g. unsteady high chair, broken cot, paint chipping).

Sufficient number of pieces of routine furniture in good repair. Seats for feeding are comfortable and supportive (e.g. foot-rest, side and back supports if needed, non-slippery surface, safety belt used if needed). S o m e child-sized furnishings used with toddlers.

\ 5

Everything in 3 plus: S ome adult furniture for use in routine care. Convenient storage for each child's possessions. Furniture suitable for individualised care of infants (e.g. feeding chairs rather than group feeding table). Routine care furnishings accessible and convenient.

6 7 Everything in 5 plus: Furnishings permit appropriate independence for toddlers (e.g. toddlers use small chairs, tables and cots when ready). Comfortable adult furniture for use in routine care. Most furnishings used with toddlers are child-sized.

Comments:

B. Personal care routines - 1. Meals and snacks.

Inadequate Minimal Adequate Good Excellent 1 2 3 4 5 6 7

Meal/snack schedule does not meet individual needs. Food service is not sanitary (e.g. care-giver does not wash o w n hands before preparing food or feeding; same sink used for meals/snacks and nappies). Food served is of questionable nutritional value or not age-appropriate. Infants not held for bottle-feeding. Infants/toddlers put to bed with bottles.

Meal/snack schedule meets children's needs (e.g. infants on individual schedules, toddlers fed lunch when hungry). Sanitary food service. Well-balanced, age-appropriate foods served for meals and snacks. Infants held while bottle-fed. Infants/toddlers not put to bed with bottles. Children who finger feed selves have hands washed. Children encouraged but not forced to eat.

Children fed separately or in very small groups. Meals/snacks are relaxed (eg. care-giver patient with messiness, slow eaters). Children encouraged to feed them selves. Menus posted for parents. Care-giver talks with children and provides a pleasant social time.

Care-giver sits with children and uses feeding time to help children learn (eg. names foods, encourages toddlers to talk and develop self-help skills). Staff cooperate with parents to establish good food habits (eg. plan together to help child give up bottle, coordinate introduction of new foods).

Comments;

295

Appendices

B. Personal care routines - 2. Nappy changing/toileting

Inadequate Minimal Adeq ... 1 2 3 -

Basic sanitary

conditions not met (e.g. care-giver doesn't use

gloves and wash own hands after nappy change or checking a nappy, potty chairs are not sanitised immediately after use). Problems with meeting toileting needs (e.g. nappies not changed often enough, children left on toilet too long, toilet training started too early). Care-giver punishes or gets angry when toileting accidents occur.

Basic sanitary

conditions are met to avoid spread of germs (e.g. nappy area disinfected or protective pad changed after each use, toilets flushed immediately after use). Care-giver uses gloves and washes hands adequately after each nappy change/check or after helping children with toileting. Nappies checked and changed as needed. Children's hands washed after toileting or, if needed, after nappy change. Caregiver handles toileting accidents calmly.

uate Good

\ 5

Nappy change done near source of hot water.

Adaptive equipment promotes self-help (e.g. steps near sink, child-sized toilet seat). Care-giver works with parents to toilet train toddlers. Pleasant tone between adult and child. Child-sized toilet or toilet seat used in place of potty-chair.

Excellent

6 7 Nappy change/toileting used as a time to talk with and relate warmly to children.

Nappy change/toileting of toddlers used to promote self-help in cleanliness and dressing skills (e.g. hand washing, using toilet paper, buttoning and fastening clothing). Parents informed about child's nappy change/toileting during the day (e.g. written record available for infants).

Comments:

C. Listening and talking -

Inadequate 1 2

Informal use of language

Minimal Adequate 3 4

Good Excellent

7

Little or no talking to infants and toddlers. Little or no response to children's attempts to communicate through gestures, sounds or words.

Talking used mainly to control child's behaviour (e.g. "come here", "take this", "don't touch"). Some social talking to children (e.g. "What a pretty baby!"). Some response to children's attempts to communicate.

Care-giver frequently responds verbally to infants'/toddlers' crying, gestures, sounds, words, and questions. Care-giver usually maintains eye contact while talking to child. Care-giver names and talks about many objects and actions for infants/toddlers. Caregiver takes part in verbal play.

Care-giver talks to each infant and toddler during play and routines about child's activities. Care-giver repeats what toddlers say, adding words and ideas when appropriate. Care-giver adds to children's understanding of language all day (e.g. gives clear directions, repeats new words often). Care-giver maintains a good balance between listening and talking (e.g. does not overwhelm child with constant talk).

Comments:

296

Appendices

D. Learning activities -1. Active physical play

Inadequate Minimal Adequate Good Excellent ... 1 2 3 -

N o outdoor or indoor space used regularly for active physical play. N o developmentally appropriate equipment/materials. Equipment/materials generally in poor repair.

Uncluttered space provided indoors for infants and toddlers to crawl and walk around much of the day. Outdoor physical play provided for infants/toddlers at least 3 times a week year-round except in very bad weather. Some developmentally appropriate toys and equipment used daily; all equipment in good repair.

\- 5

Convenient outdoor area where infants/toddlers are separated from older children used for at least one hour daily year-round, except in very bad weather. All toys and equipment for physical activity used both indoors and outdoors are developmentally appropriate. Materials used daily stimulate variety of large muscle skills (e.g. crawling, walking, balancing, climbing, ball play). Active play areas are not crowded.

5 7

Physical play equipment changed or rotated weekly to provide new challenges either indoors or outdoors (e.g. crawling tunnel, games with bean bags, tumbling on mat, ball games). Care-giver talks to children about their activities (e.g. explains safety rules, names up/down, in/out).

Comments:

D. Learning activities - 2. Pretend play

Inadequate Minimal Adequate Good Excellent 1 2 3 4 5 6 7

No materials available for pretend play.

Enough pretend-play materials accessible for number of children present. Materials are developmentally appropriate (e.g. easy to open purses, hats, clothes shortened and easy to put on, no necklaces with small beads).

Variety of developmentally appropriate pretend-play materials accessible daily, indoors. Props focus on what toddlers see in real life (e.g. household routines, parents' work). Some child-sized play furniture for toddlers (e.g. small stove, baby stroller).

Materials accessible to children for both indoor and outdoor use. Materials well organised for independent use (e.g. play dishes in separate box or on shelf instead of piled in a toy chest; play clothes hung on pegs). Care-giver pretends with children in play (e.g. talks to child on toy telephone).

Comments:

297

Appendices

D. Learning activities - 3. Cultural awareness

Inadequate Minimal Adequate Good Excellent 1 _

All dolls, books and visible pictures are of one race only.

3 -

Some evidence of ethnic and racial variety in toys (e.g. multiracial or multicultural dolls). S ome evidence of ethnic and racial variety in books and pictures.

1 5

Cultural awareness evidenced by generous inclusion of multiracial books and dolls for children to use. Pictures on child's eye level that show people of various ages from infancy through old age.

5 7

Non-sexist pictures displayed of men and women, boys and girls in similar work and play roles. Cultural awareness shown in a variety of activities (e.g. various types of music, celebration of different holidays and customs, ethnic foods served).

Comments:

E. Interaction -1. Peer interaction

Inadequate Minimal Adequate Good Excellent

... 1 2 3

Little or no developmentally appropriate peer interaction possible (e.g. children separated in cribs or playpens while awake). Little or no care-giver guidance to ensure positive peer interactions.

Children allowed to move freely so natural groupings and interactions can occur much of the day. Non-mobile infants taken out of cribs, playpens and swings for some supervised playtime near others.

1 5

Peer interaction is usually positive (e.g. infants watch and react to others; toddlers play side by side with few conflicts). Care-giver models positive social interaction (e.g. is warm and affectionate, demonstrates gentle touching, helps toddler walk around non-mobile baby).

5 7

Care-giver reinforces positive social interaction (e.g. smiles and talks to babies who notice other children, praises child for getting his or her own duplicate

toy). Care-giver points out and talks about instances of positive social interaction (e.g. helps children notice sharing, turn taking, comforting).

Comments:

298

Appendices

E. Interaction - 2. Caregiver-child interaction.

Inadequate Minimal Adequate Good Excellent 1 _

Little affection shown. Care is impersonal (e.g. care-giver does not often respond to, smile at, talk to, or listen to children). Uneven amount of attention given to each child - care-giver shows favouritism. Physical contact used principally for routines or controlling children's behaviour (e.g. changing baby's nappy, picking up toddler to stop what he or she is doing).

1 3 -

Some smiling, talking and affection shown to all children. Care-giver shows warmth in physical contact during routines (e.g. holds child gently, smiles while feeding child). Care-giver responds sympathetically to help children who are hurt or upset. Children happy most of the time.

1 5

Frequent positive caregiver-child interaction throughout the day (e.g. care-giver initiates verbal and physical play, responds when child initiates interactions, shows delight in child's activity). Care-giver and children relaxed, voices pleasant, frequent smiling. M u c h holding, patting, and physical warmth shown throughout the day. Care-giver-child interaction is consistent across all care-givers, including relief staff.

5 7

Responsibility for a small number of children is assigned to one primary care-giver. Care-giver varies interactive style to meet children's individual needs (e.g. calmer with timid baby, more active with outgoing baby). Care-giver is sensitive about children's feelings and reactions (e.g. avoids abrupt interruptions, warns baby before picking him or her up).

Comments:

E. Interaction - 3. Discipline


... 1 2 3

Discipline is EITHER so strict that children are punished often O R so lax that there is no control. Children controlled with severe methods such as spanking, shouting, confining children for long periods or withholding food.

Care-giver never uses physical punishment or severe discipline. Care-giver usually maintains enough control to prevent children from hurting one another.

•

1 5

F e w discipline problems because the program is set up to avoid conflict and promote developmentally appropriate interaction (e.g. duplicate toys accessible, child with favourite toy given protected place to play, smooth transitions). Alternative methods of discipline used effectively (e.g. removing child from negative activity, redirection). Expectations are realistic and based on age and ability of each child.

6 7

Attention frequently given for good behaviour. Care-giver assists children to avoid or solve problems (e.g. helps children avoid serious conflict over toy, stays close to toddler who bites). Rules are simple and explained to toddlers.

Comments:

299

Appendices

D . Program structure - Supervision of daily activities

Inadequate Minimal Adequate Good Excellent 1 2 3

Insufficient care-giver supervision provided to protect health and safety, and satisfy routine-care needs (e.g. care-giver leaves children and cannot see, hear, or reach them, even for a short time).

Care-giver is within sight, hearing and easy reach of children at all times. Attention is on caregiving not on other tasks or interests. Sufficient supervision to meet every child's routine-care needs with little crying, waiting or regimentation.

1 5

Care-giver maintains supervision of whole group while working with one child or a small group. Care-giver watches carefully and intervenes to avoid problems (e.g. avoids conflict by bringing out duplicate toy, moves active play that is disrupting quiet play). Care-giver plays with children and shows appreciation for what they do. Care-giver reacts quickly to solve problems in a comforting and supportive way.

5 7

Supervision is provided with each child's learning needs in mind (e.g. care-giver moves non-mobile infants to avoid boredom, does planned activity with a small group while others are sleeping). Balance is maintained between child's needs for independent exploration and caregiver support.

Comments:

Adult needs - Provisions for parents


... 1 2 3

No information concerning program given to parents in writing. N o possibility for parent involvement in children's daily program (e.g. parents discouraged from observing or participating in program).

Parents given administrative information about program in writing (e.g. fees, hours of service, health rules for attendance). S o m e child-rearing information available (e.g. parent conferences upon request, some parenting materials). S o m e possibilities for parent involvement in children's' program (e.g. m a y bring special birthday treat, eat lunch with child).

1 5 e

Prior to enrolment, parents urged to observe in child's group. Parents made aware of philosophy and approaches practiced (e.g. parent handbook, discipline policy, descriptions of activities). Appropriate parenting information regularly available (e.g. periodic conferences, parent educational meetings, newsletters, health care and development leaflets). Variety of alternatives used to encourage parent involvement in children's program.

, 7

Parents asked for an evaluation of the program annually (e.g. parent questionnaires, group evaluation meetings). Parents referred to other professionals for special parenting help. Parents involved in decision-making roles in program along with staff (e.g. parent representative on centre committee or management board).

Comments:

300

Appendices

Appendix Fi.

Play materials likely to be found in the outdoor environment in L D C C C s in Sydney.

*

W i l l 1 ̂ M ^W^*"-:~>?~''~ "***•'"'*m

^wjp*"5 » ": 'Dress-ups' - clothes, shoes and mirror

set up in sandpit.

Buckets, spades and scoops in sandpit.

'Home corner' set up in sandpit.

301

Appendix Fii.

Play materials provided by researcher

for dyad play.

Appendices

Plastic tea-.set, cutlery .set, coffee pot,

two dolls. (These materials are likely to

elicit emerging representational skills.)

Table blocks, small, plastic wheeled-

vehicles and human-like figurines.

(These materials are likely to elicit "role

and situational transformations"

[Westby, 1991, p. 147].)

Play-dough and accessories. (Likely to

elicit emerging representational skills.)

302

Appendices

Appendix Gi Play Coding sheet 1. name

sex

20sec

1 30s

2 lm

3 1.30s

4 2m

5 2.30s

6 3m

7 3.30s

8 4m

9 4.30s

10.5m

115.30

12 6m

13 6.30

14 7m

15 7.30

16 8m

17 8.30

18 9m

19 9.30

20 10m

totals

age

cat soc'l play

a b c d e f

date of ob

type

of play

1 2 3 4 5

LDCC

peer S/A/P

O/P

setting m/a s/a dyad

imitat'n

session am mid pm

pretend play

1/D 1/T I/O 1/S 2/T 2/0 2/S 3/0 4/T 4/Oi 4/Oii

material equip't

O/P comm'n

1 2 3 4 5 6 7

level of ply H/M/L

H

M

L

them'ic unit B2-3m L4m+

B

L

other info

303

Appendices

Appendix Gii Symbolic play coding sheet (in accordance with Westby's [1991] Scale for Assessing Children's Pretend Play).

Pretend play coding sheet Time 30 sees

1.30s 2. Im 3.1.30s

4. 2m 5.2.30s

6. 3m 7.3.30s

8. 4m 9.4.30s

10.5m 11.5.30s

12.6m 13.6.30s

14.7m 15.7.30s

16.8m 17.8.30s

18.9m 19.9.30s

20.10m

totals

Sy'ic 1 1/D pret'ds uses life like props

decon-textual -isat'n

1/D

Sy'ic 1 1/T famil'r e'day act'ies

themes, scheme scripts

1/T

Sy'ic 1 I/O single pretend actions

organ'n coher'e logic

I/O

Sy'ic 1 1/S self as agent

self-other rel'ns

1/S

Sy'ic 2 2/T act'ies of famil'r others


2/T

Sy'ic 2 2/0 two actions comb'd


2/Q

Sy'ic 2 2/S ch. acts on 1 or more objects or people

self-other rel'ris

2/S

Sy'ic 3 3/0 elab'ed single scheme


3/Q

Sy*ic4 4/T repres'ts less freq'ntly perso'n' y exper'ed events


4/T

Sy'ic4 4/0 evolv'g episode seque'es


4/0

Sy'ic4 4/S self-other rel'ns

self-other rel'ns

4/S

other info

304

Appendices

Appendix Giii Individual S u m m a r y Sheet for Levels and Dimensions of Symbolic Play in each of the Three Conditions.

Participant Id: Age: LDCCC Same-age Mixed-age Dyad Condition One Condition Two Condition Three

Min 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Total

Sec 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 9 9.5 10

LSP D T 0 s LSP D T 0 D T O LSP D T 0 S TS

Same-age Mixed-age Dyad Condition One Condition Two Condition Three

Min 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Total

Sec 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 9 9.5 10

LSP D T O S LSP D T O D T O LSP D T O S B

305

Appendices

Same-age Mixed-age Dyad Condition One Condition T w o Condition Three

Min 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Total

Sec 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 9 9.5 10

LSP D T 0 s LSP D T 0 D T 0 LSP D T 0 s PD

Key: - LSP denotes level of symbolic play according to Westby (1991) - D denotes dimension one of symbolic play - 'decontextualisation' - T denotes dimension two of symbolic play - 'thematic content' - O denotes dimension three of symbolic play - 'organisation of themes' - S denotes dimension four of symbolic play - 'self-other relationships' - T denotes 'tea-set' play materials (supplied for first session of dyad condition) - B denotes 'blocks' play materials (supplied for second session of dyad condition) - P D denotes 'play-dough' play materials (supplied for third session of dyad condition)

306

Appendices

Appendix Giv

Behaviour codes - identifying amount and form of information given to younger participant by older peer.

Adapted from: Beizer, L., & Howes, C. (1992). In C. Howes, O. Unger & C. C. Matheson (Eds.), The

collaborative construction of pretend: social pretend play functions. (25-43). N e w York: State University of N e w York Press.

A. - behaviour codes - reflecting the older child's responses to the younger child.

a. verbalising - narrating the younger child's pretend play, labelling objects as the

younger child uses them in pretend play

b. indirect support - physically assisting the younger child with props or refocusing the

child's attention on props

- elaborating on the younger child's pretend play - by extending the child's play, or by

joining in and enacting similar or complementary pretend play

c. positive support - showing younger child encouragement, nodding, smiling,

comments indicating acceptance of play

d. instruction - demonstrating literal (not symbolic) use of a prop to younger child

e. clarifying - repeating the younger child's comment or asking a question to clarify

intent or statement

iii. direct pretend suggestions

a. direct verbal suggestions, instructions or invitations to engage in pretend play

activities

b. gestures of use of modelling of specific acts of pretend play

c. a combination of a and b

307

Appendices

Appendix Gv

Older-Peer Communicative Behaviours by Condition and Type of

Communicative Behaviour

Condition 2 Condition 3 Mixed-age play. Dyad play

Communicative Indirect Direct Indirect Direct Behaviours

Participant

Id 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

0 0 0 1 0 0 0 11 0 0 0 0 0 0 8 0 0 0 0 0 0 1 0 0

0 0 0 8 19 9 9 0 9 9 9 9 9 9 9 9 9 9 9 9 6 7 0 0

1 0 8 1 1 2 2 4 0 0 1 2 2 1 2 7 0 0 1 1 20 0 6 1

18 21 9 38 49 36 19 23 43 42 35 33 23 43 17 14 26 45 46 47 26 40 4 19

Total 21 175 63 716

308

Appendices

Appendix Gvi

Frequency of Each Toddler Participant's Imitative Behaviour in Condition T w o

and Condition Three

Condition Two Condition Three

Mixed-age play Dyad play

Participant Id N imitative

events

N imitative

events

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

0

1

0

0

1

0

0

0

0

0

0

1

0

0

0

1

0

0

0

12

2

8

1

0

0

0

2

2

4

2

5

3

0

17

0

3

5

1

3

10

8

3

4

6

15

3

14

6

Total 27 116

309

Appendices

Appendix Hi Extracts from Minitab Output for Friedman's Two-way Analysis of Variance by Ranks

Highest score in Dimension One 'decontextualisation' compared for the three conditions (same-age free play, mixed-age free play and dyad free play).

MTB > Friedman c50 c48 c49.

Friedman test of h dec by SMD blocked by child

S = 2.44 df. = 2 p = 0.296 S = 4.59 df. = 2 p = 0.101 (adjusted for ties)

Est. Sum of S M D N Median R A N K S

1 24 1.0000 46.5 2 24 1.0000 43.5 3 24 1.0000 54.0

Grand median = 1.0000


Highest score in Dimension Two 'thematic content' compared for the three conditions (same-age free

play, mixed-age free play and dyad free play).

Friedman test of h the by SMD blocked by child

S=1.90 df. = 2 p = 0.388 S = 3.50 df. = 2 p = 0.174 (adjusted for ties)


1 24 2.0000 47.5 2 24 2.0000 43.5 3 24 2.0000 53.0

Grand median =2.0000


Highest score in Dimension Three 'organisation of themes' compared for the three conditions (same-

age free play, mixed-age free play and dyad free play).

Friedman test of h org by SMD blocked by child


Est. Sum of

S M D N Median R A N K S 1 24 1.3333 44.5 2 24 1.1667 42.0 3 24 2.0000 57.5

Grand median = 1.5000 M T B > Friedman c53 c48 c49.

310

Appendices

Highest score in Dimension Four 'self-other relationships' compared for the three conditions (same-age free play, mixed-age free play and dyad free play).

Friedman test of h sel by SMD blocked by child


Est. Sum of SMD N Median RANKS

1 24 2.0000 45.5 2 24 2.0000 42.5 3 24 2.0000 v56.0



Most commonly occurring score in Dimension One 'decontextualisation' compared for the three conditions (same-age free play, mixed-age free play and dyad free play).

Friedman test of mc dec by SMD blocked by child


Est. Sum of SMD N Median RANKS

1 24 1.0000 47.0 2 24 1.0000 45.5 3 24 1.0000 51.5

Grand median = 1.0000


Most commonly occurring score in Dimension Two 'thematic content' compared for the three

conditions (same-age free play, mixed-age free play and dyad free play).

Friedman test of mc the by SMD blocked by child


Est. Sum of

S M D N Median RANKS 1 24 1.5000 48.0 2 24 1.5000 47.5 3 24 1.5000 48.5

Grand median = 1.5000 MTB > Friedman c56 c48 c49.

311

Appendices

^ Z T ^ ^ T ^ SC°re injDimension Three 'organisation of themes' compared for the three

conditions (same-age free play, mixed-age free play and dyad free play).

Friedman test of mc org by SMD blocked by child



1 24 1.6667 45.5 2 24 1.8333 46.5 3 24 2.0000 52.0



Most commonly occurring score in Dimension Four 'self-other relationships' compared for the three conditions (same-age free play, mixed-age free play and dyad free play).

Friedman test of mc sel by SMD blocked by child



1 24 2.0000 46.5 2 24 2.0000 45.0 3 24 2.0000 52.5



Frequency of events of symbolic play in any dimension compared for the three conditions (same-age free play, mixed-age free play and dyad free play).

Friedman test of hap by SMD blocked by child



1 24 7.833 39.0 2 24 7.167 40.5 3 24 20.500 64.5



312

Appendices

Frequency of non-events of symbolic play in any dimension compared for the three conditions (same-age free play, mixed-age free play and dyad free play).

Friedman test of non hap by SMD blocked by child



1 24 49.58 58.0 2 24 49.42 61.0 3 24 28.25 25.0

Grand median =42.42

313

Appendix Hii

Extracts from 20 page output for Kruskal-Wallis One-Way Analysis of Variance by Ranks statistical test.

MTB > Code (17:19) 1 (20:22) 2 (23:26) 3 MTB > Kruskal-Wallis 'HSAMED' 'age_gp'.

;27:31) 4 'AGE' 'BLANK1

age_gp 1 2 3 4 Overall

H = 6.65 H = 8.94

N 5 6 6 7

24

DF = DF =

1. 0. 1 1

3 3

Median .OOE+00 .00E+00 .00E+00 .OOE+00

P = 0. P = 0.

084 030

Ave Rank 13.0 6.3 16.2 14.3 12.5

(adjusted f

z 0.18

-2.47 1.47 0.79

or tie

MTB > Kruskal-Wallis 'HSAMET


N 5 6 6 7

24

2 0 2 2

Median .00E+00 .OOE+00 .OOE+00 .00E+00

'age_gp'.

Ave Rank 14.5 5.5 16.1 14.0

Z 0.71 -2.80 1.43 0.67

12.5

H = 8.14 D F = 3 P = 0.043 H = 10.26 DF = 3 P = 0.016 (adjusted for ties]

MTB > Kruskal-Wallis 'HSAMEO


H = 6.77 H = 7.33

N 5 6 6 7

24

DF = DF =

2 0 2 2.

3 3

Median .00E+00 .OOE+00 .OOE+00 •00E+00

P = 0 P = 0

'age_gp'.

»/e Rank 14.5 6.1

15.6 13.9 12.5

Z 0.71 -2.57 1.23 0.64

P = 0.062 (adjusted for ties)

MTB > Kruskal-Wallis 'HSAMES'


H = 5.56 H = 8.73

N 5 6 6 7

24

DF = DF =

2 0 2 2

3 3

Median .OOE+00 .OOE+00 .00E+00 .00E+00

P = 0 P = 0

'age_gp'.

Ave Rank 14.0 6.7

14.0 15.1 12.5

Z 0.53 -2.33 0.60 1.17


314

MTB > Kruskal-Wallis 'HMIXEDD' 'age_gp'.


H = 2.84 H = 3.84

N 5 6 6 7

24

DF = DF =

3 3

Median 1.0000 0.5000 1.0000 1.0000

P = 0 P = 0

e Rank 11.9 8.8 13.8 15.1 12.5

Z -0.21 -1.50 0.50 1.14

0.417 P =: 0.280 (adjusted for ties)

MTB > Kruskal-Wallis 'HMIXEDT


H = 2.13 H = 2.70

N 5 6 6 7

24

DF = DF =

3 3

Median 2.000 1.000 2.000 2.000

P = 0 P = 0

'' 'age_ _gp'

Ave Rank 11.5 9.5 13.5 14.9 12.5

Z -0.36 -1.20 0.40 1.08


MTB > Kruskal-Wallis 'HMIXEDO' 'age_gp'


H = 2.65 H = 3.10

N 5 6 6 7 24

DF = DF =

MTB > Kruskal-


H = 2.92 H = 3.54

N 5 6 6 7

24

DF = DF =

3 3

Median 2.0000 0.5000 2.0000 2.0000

P = 0. P = 0.

448 377

Ave Rank 12.3 8.7 14.3 14.4 12.5

(adjusted

Wallis 'HMIXEDS' ' age_gr.

3 3

Median 2.000 1.000 2.000 2.000

P = 0. P = 0.

404 315

Ave Rank 11.2 9.3 13.1 15.7 12.5

(adjusted

z -0.07 -1.53 0.73 0.83

for ties

i 1 .

Z -0.46 -1.30 0.23 1.43

for ties

315

MTB > Kruskal-Wallis 'HDYADD' 'age_gp'.


N 5 6 6 7 24

Mec 1. 1. 1. 1.

lian .000 .000 .000 .000

Ave Rank 10.5 10.5 14.5 13.9 12.5

Z -0.71 -0.80 0.80 0.64

H = 1.65 DF - 3 P = 0.649 H = 3.94 DF = 3 P = 0.268 (adjusted for ties]

MTB > Kruskal-Wallis 'HDYADT' 'age_gp'.


N 5 6 6 7

24

Median 2.000 2.000 2.000 2.000

Ave Rank 12.5 12.5 12.5 12.5 12.5

Z 0.00 0.00 0.00 0.00

H = 0.00 DF = 3 P = 1.000

* NOTE * All values in column are identical

MTB > Kruskal-Wallis 'HDYADO' 'age_gp'.


N 5 6 6 7 24

Median 000 000 000 000

Ave Rank 11.0 11.0 15.1 12.6 12.5

Z -0.53 -0.60 1.03 0.06

H = 1.30 DF = 3 P = 0.729 H = 3 . 9 4 DF = 3 P = 0.269 (adjusted for ties)

MTB > Kruskal-Wallis 'HDYADS' 'age_gp'.


N 5 6 6 7

24

Median 000 000 000 000

Ave Rank 11.7 15.5 11.8 11.1 12.5

Z -0.28 1.20 -0.27 -0.64

H = 1.48 DF = 3 P = 0.686 H = 2 . 5 8 D F = 3 P = 0.460 (adjusted for ties)

316

MTB > Kruskal-Wallis 'MCSAMED' 'age_gp'.


H = 5.68 H = 8.93

N 5 6 6 7 24

DF = DF =

MTB > Kruskal-


N 5 6 6 7 24

Median 1.OOE+00 0.OOE+00 1.OOE+00 1.OOE+00

3 P = 0. 3 P = 0.

128 030

Ave Rank 14.0 6.7 15.6 13.8 12.5

(adjusted

Wallis 'MCSAMET' 'age_gp

Median 2.OOE+00 0.OOE+00 1.50E+00 2.OOE+00

Ave Rank 14.2 6.4 13.3 15.9 12.5

Z 0.53 -2.33 1.23 0.57

for ties

r

Z 0.60 -2.43 0.30 1.49

H = 6.38 DF = 3 P = 0.095 H = 7 . 8 0 D F = 3 P = 0.050 (adjusted for ties)

MTB > Kruskal-Wallis 'MCSAMEO' 'age_gp'.


H = 6.24 H = 6.92

N 5 6 6 7

24

DF = DF =

2 0. 1. 1

3 3

Median .OOE+00 .OOE+00 .50E+00 •00E+00

P = 0. P = 0.

100 074

Ave Rank 16.3 6.6 14.7 13.0 12.5

(adjusted

Z 1.35 -2.37 0.87 0.22

for tie

MTB > Kruskal-Wallis 'MCSAMES' 'age_gp'


H = 4.82 H = 7.58

N 5 6 6 7

24

DF = DF =

2 0 2 2

3 3

Median .00E+00 .OOE+00 .00E+00 .OOE+00

P = 0. P = 0.

185 056

Ave Rank 16.0 7.3 12.8 14.1 12.5

(adjusted

z 1.24 -2.07 0.13 0.73

for ties)

317

MTB > Kruskal-Wallis 'MCMIXEDD


N 5 6 6 7 24

Median 1.0000 0.5000 1.0000 1.0000

'D ' ' age_gp '

Ave Rank 12.3 9.0

14.2 14.2 12.5

Z -0.07 -1.40 0.67 0.76

H = 2 . 2 2 DF = 3 P = 0.528 H = 3.22 DF = 3 P = 0.359 (adjusted for ties)

MTB > Kruskal-Wallis 'MCMIXEDT' 'age_gp'.


N 5 6 6 7

24

Median 1.000 1.000 2.000 2.000

Ave Rank 9.6

10.5 13.8 15.2 12.5

Z -1.03 -0.80 0.50 1.21

H = 2 . 5 4 D F = 3 P = 0.468 H = 2.98 D F = 3 P = 0.394 (adjusted for ties)

MTB > Kruskal-Wallis 'MCMIXEDO' 'age_gp'.


N 5 6 6 7

24

Median 1.0000 0.5000

0000 0000

Ave Rank 11.7 9.3

14.2 14.4 12.5

Z -0.28 -1.27 0.67 0.83

H = 2 . 0 8 DF = 3 P = 0.555 H = 2 . 3 8 D F = 3 P = 0.497 (adjusted for ties)

MTB > Kruskal-

age_gp N 1 5 2 6 3 6 4 7 Overall 24

-Wallis 'MCMIXEDS

Median 2.000 1.000 2.000 2.000

' 'age_gp'

Ave Rank 11.8 9.8

13.8 14.2 12.5

•

Z -0.25 -1.10 0.53 0.76

H = 1.58 DF = 3 P = 0.664 H = 2.13 D F = 3 P = 0.545 (adjusted for ties)

MTB > Kruskal-

age_gp N 1 5 2 6 3 6 4 7 Overall 24

-Wallis 'MCDYADD ' 'age_ .gp'

Median Ave Rank 1.000 1.000 1.000 1.000

12.5 12.5 12.5 12.5 12.5

z 0.00 0.00 0.00 0.00

H = 0.00 DF = 3 P = 1.000

* NOTE * All values in column are identical

318

MTB > Kruskal-Wallis 'MCDYADT' 'age_gp'


N 5 6 6 7 24

Median 1.000 1.000 1.000 2.000

Ave Rank 11.8 9.0 11.0 17.3 12.5

Z -0.25 -1.40 -0.60 2.13

H = 5 . 0 0 DF = 3 P = 0.172 H = 6.70 DF = 3 P = 0.082 (adjusted for ties)

MTB > Kruskal-Wallis 'MCDYADO' 'age_gp'.


N 5 6 6 7 24

Median 2, 1, 1 2

000 000 500 000

Ave Rank 12.7 9.5 11.5 15.8 12.5

Z 0.07 -1.20 -0.40 1.46

H = 2.72 DF = 3 P = 0.438 H = 3.72 D F = 3 P = 0.294 (adjusted for ties)

MTB > Kruskal-Wallis 'MCDYADS' 'age_gp'.


H = 1.42 H = 4.31

N 5 6 6 7 24

DF = DF =

MTB > Kruskal-


N 5 6 6 7

24

: 3 : 3

Median 2.000 2.000 2.000 2.000

P = 0. P = 0.

Wallis 'N_

1 0 9 7

Median .10E+01 .00E+00 .50E+00 .00E+00

702 230

.SAME

Ave Rank 11.6 14.0 10.0 14.0 12.5

(adjusted

' 'age_gp'

Ave Rank 16.2 5.5 15.5 13.3 12.5

Z -0.32 0.60 -1.00 0.67

for ties

Z 1.32 -2.80 1.20 0.35

H = 8.42 D F = 3 P = 0.038 H = 8.51 DF = 3 P = 0.037 (adjusted for ties)

MTB > Kruskal-


H = 1.96 H = 2.00

N 5 6 6 7 24

DF = DF =

•Wallis 'N_

= 3 = 3

Median 11.000 1.000 5.000 4.000

P - 0. P = 0.

.MIXED' ' age_gp

581 573

Ave Rank 15.3 9.5

12.3 13.3 12.5

(adjusted

I

Z 1.00 -1.20 -0.10 0.35

for ties

319

Appendices

MTB > Kruskal-Wallis 'N_DYAD' 'age_gp'

age_gp N Median 1 5 12.00 2 6 15.50 3 6 27.50 4 7 23.00 Overall 24 12.5

H = 7.93 DF=3 P = 0.047 H = 7.96 D F = 3 P = 0.047 (adjusted for ties)

MTB >

Ave Rank 7.3 8.8 16.8 15.7

Z -1.85 -1.47 1.70 1.43

320

Appendices

Appendix Hiii

Explanation of Logistic Regression Analysis

We are interested in the odds that a child showed symbolic play at a level that was greater than expected by Westby's scale, and whether these are the same under different conditions after adjusting for any relevant covariates.

To illustrate the use of the above analysis, we introduce the following notation:

The response observed is one (1) when the child shows symbolic play at a level that is greater than expected by Westby's scale, and zero (0) when less than or equal to that expected. That is, the response is one (1) when play is complex and zero (0) otherwise.

In the logistic regression model resulting from the above analysis there are two predictor variables: - Condition with two levels: Condition 3 (dyad free play) and Conditions 1 or 2 (same-age free play or mixed-age free play) denoted by xc1 and xc° respectively.

- Materials with two levels: satisfactory provision of play materials and unsatisfactory provision of play materials, denoted by X M 1 and X M ° respectively.

Then the odds ratio of complex play under a combination of xc and XM relative to another combination is given by

O R = exp(-0.8061(xc1 - xc°) + 1.6698(xM

1 - xM°)).

We chose xc° = 0 and XM° = 0 as the referent category.

When xc1 = 0 and XM1 = 1 OR = exp(-0.8061(0-0) + 1.6698(1-0)) = 5.31;

when xc1 = 1 and XM1 = 0 OR = exp(-0.8061) = 0.45; and

when xc1 = 1 and xMx = 1 OR = exp(-0.8061 + 1.6698) = 2.34.

Since (0.45 x 5.31) = 2.39 is approximately equal to 2.34, we can say that the variables Condition and Materials act independently of each other. W h e n the two variables are fitted separately, the respective

odds ratios are 0.50 and 5.13, which confirms their independence.

Thus: - Children are 2.24 times as likely to display complex play when in Condition 3 (dyad free play) than when in Conditions 1 or 2 - same-age free play or mixed-age free play - (95% confidence interval is (1.01,4.96)).

- Children with unsatisfactory Materials are 5.31 times as likely to display complex play than those with satisfactory Materials (95% confidence interval

is (1.05,25.6)).

321

Appendices

Appendix Hiv

Extract from SAS output for Logistic Regression Analysis

The GENMOD Procedure

Model Information

Description

Data Set Distribution Link Function Dependent Variable Observations Used Number Of Events Number Of Trials

Value

NAN.SIBBA BINOMIAL LOGIT HS 72 20 72

Class Level Information

Class

CHILD

MATPROV DYAD

Levels

24

2 2

Values

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 0 1 0 1

Parameter Information

Parameter Effect MATPROV DYAD

PRM1 PRM2 PRM3 PRM4 PRM5

INTERCEPT DYAD DYAD MATPROV MATPROV

0 1

0 1

Criteria For Assessing Goodness Of Fit

Criterion DF Value Value/DF

Deviance 69 74.6985 1.0826 Scaled Deviance 69 74.6985 1.0826 Pearson Chi-Square 69 70.8012 1.0261 Scaled Pearson X2 69 70.8012 1.0261 Log Likelihood -37.3492

Analysis Of Initial Parameter Estimates

Parameter DF Estimate Std Err Chi Square Pr>Chi

INTERCEPT 1 -1.0480 0.4926 4.5261 0.0334 DYAD 0 1 -0.8102 0.5894 1.8898 0.1692 DYAD 1 0 0.0000 0.0000

MATPROV 0 1 1.6902 0.5854 8.3373 0.0039 MATPROV 1 0 0.0000 0.0000 SCALE 0 1-0000 0.0000

NOTE: The scale parameter was held fixed.

322

LR Statistics For Type 1 Analysis

Source Deviance DF ChiSquare Pr>Chi

INTERCEPT DYAD MATPROV

85.0813 83.4287 74.6985

0 1 1

1.6526 0.1986 8.7303 0.0031

LR Statistics For Type 3 Analysis

Source DF ChiSquare Pr>Chi

DYAD MATPROV

1.8980 0.1683 8.7303 0.0031

GEE Model Information

Description Value

Correlation Structure Exchangeable Subject Effect CHILD (24 levels) Number of Clusters 24 Correlation Matrix Dimension 3 Maximum Cluster Size 3 Minimum Cluster Size 3

Analysis Of GEE Parameter Estimates Empirical Standard Error Estimates

Empirical %% Confidence Limits Parameter

INTERCEPT DYAD DYAD MATPROV MATPROV Scale

0 1 0 1

Estimate

-1.0221 -0.8061 0.0000 1.6698 0.0000 0.9855

Std Err

0.5292 0.4061 0.0000 0.8033 0.0000

Lower

-2.0594 -1.6020 0.0000 0.0953 0.0000

Upper

0.0152 -0.0101 0.0000 3.2443 0.0000

Z

-1.931 -1.985 0.0000 2.0786 0.0000

Pr>|Z|

0.0534 0.0472 0.0000 0.0376 0.0000

323

Appendices

Appendix I

SUMMARY OF STATISTICAL RESULTS

Key to the three conditions under which children were observed and video-taped and the four

dimensions - or behaviours - of symbolic activity:

- Condition One - Same-age play;

- Condition T w o - Mixed-age play;

- Condition Three - Dyad play;

- Dimension One - 'Decontextualisation';

- Dimension T w o - 'Thematic content';

- Dimension Three - 'Organisation of themes';

- Dimension Four - 'Self-other relationships'.

Results using Friedman's Rank Test

The Frequency of Symbolic-Play Events:

- Significant differences between the means for the frequency of

symbolic-play events were found in Condition Three (p = .000).

The Highest Levels of Symbolic Play:

- Significant differences between the means for the participants'

highest levels of Dimension Three (p = .014), and Dimension Four

(p = .021) were found in Condition Three.

The Most-Frequently Occurring Levels of Symbolic Play:

- N o significant differences between the means for the participants'

most-frequently exhibited levels in any of the four dimensions of

symbolic-play behaviour were found between: Condition One and

Condition Two; Condition One and Condition Three; and Condition

T w o and Condition Three.

Results using the Kruskal-Wallis Test

The Relationship Between Participants' Birth Order and the Frequency of Symbolic-Play Events:

- N o significant differences between the means for the frequency of

symbolic-play events were found in any of the four dimensions in any

condition. However, in Condition Two, for first-born participants a

p value of .061 in Dimension One approaches significance.

The Relationship Between Participants' Birth Order and the Highest Levels of Symbolic Play:


each of the four symbolic-play behaviours were found for first-born

participants in Condition T w o only: Dimension One (p = .053);

Dimension T w o (p = .044); Dimension Three (p = .029); and

Dimension Four (p = .036).

324

Appendices

The Relationship Between Participants' Birth Order and the Most-Frequently Occurring Levels of

Symbolic Play:

- Significant differences between the means of the most-frequently

occurring levels of symbolic-play behaviours in three dimensions were

found for first-born participants in Condition T w o only: Dimension

T w o (p =.053); Dimension Three (p = .021); and Dimension Four

(p = .014).

The Relationship Between Participants' Age Group and the Frequency of Symbolic-Play Events:

Though unanticipated earlier in the project, the following two results (using the Kruskal-Wallis

test) have proved useful for discussion in relation to the major hypothesis and additional questions:

- Significant differences between the means of the frequency of

symbolic-play events in all four dimensions were found for

participants aged 20 to 22 months in Condition One (p = .037) and

Condition Three (p = .047).

The Relationship Between Participants' Age Group and the Highest Levels of Symbolic-Play:


each of the four symbolic-play behaviours were found for participants

aged 20 to 22 months in Condition One only: Dimension One

(p = .030); Dimension T w o (p = .016); and Dimension Four (p = .033).

A p value of .062 in Dimension Three approaches significance.

The Relationship Between Participants' Age Group and the Most-Frequently Occurring Levels of

Symbolic-Play:

- Significant differences between the means of the most-frequently-

occurring levels of symbolic play were found for participants aged 20 to

22 months in Condition One: Dimension One (p = .030) and, Dimension T w o (p =

.050). A p value of .074 for Dimension Three and a p value of .056 for Dimension Four

approach significance. (There were no significant differences for the other age groups.)

Results using Logistic Regression Analysis

The Relationship Between Individual Age and the Complexity of Symbolic Play

A relationship between age and the complexity of symbolic play (symbolic activity that is either

'greater than' or 'less than or equal to' that expected) was found in:

- Condition One in Dimensions T w o (p = .003) and Four (p = .002)

and the p value of .062 for Dimension One approaches significance;

- Condition T w o in Dimension One (p = .049);

- Condition Three in three dimensions - Dimension One (p = .030),

Dimension Three (p = .000) and Dimension Four (p = .000).

325

Appendices

Variables Useful to Explain the Likelihood of Participants' Showing Complexity of Symbolic Play


ratio) of participants' showing complexity of symbolic play in

Dimension T w o were Condition Three - dyad play - (likelihood 2.2,

p = .048) and unsatisfactory provision of play materials (likelihood

5.5, p = .034).


ratio) of participants showing complexity of symbolic play in

Dimension Three are a weekly attendance at L D C C C of four or more

days (likelihood 14, p = .066) and unsatisfactory provision of play

materials (likelihood 12.5, p = .004).

-Variables identified as useful to explain the likelihood (the odds ratio)

of participants showing complexity of symbolic play in Dimension Four

are Condition Three -dyad play (likelihood 2.2, p = .038) and

unsatisfactory provision of play materials (likelihood 5.3, p = .047).

326

2000 toddlers' emerging symbolic play: the influence of

Documents